Re: 자가포식 유도체 '스퍼미딘' 약물 역동학 15mg/day이상 사용해야

[문형철]

https://kr.iherb.com/pr/force-factor-spermidine-120-vegetable-capsules-5-mg-per-capsule/143321?_gl=1*7lwyfi*_up*MQ..&gclid=Cj0KCQjwqc6aBhC4ARIsAN06NmOGhfHAyla-xkmQmkwUTeApBwqamFsqN4L-MLo-GOLG18nbXpr2DfYaAjm5EALw_wcB&gclsrc=aw.ds

High-Dose Spermidine Supplementation Does Not Increase Spermidine Levels in Blood Plasma and Saliva of Healthy Adults: A Randomized Placebo-Controlled Pharmacokinetic and Metabolomic Study

by 

Stefan Senekowitsch

Department of Biopharmaceutics and Pharmaceutical Technology, University of Greifswald, 17489 Greifswald, Germany

2

Institute of Nutritional Medicine, University Hospital Schleswig-Holstein, Campus Lübeck, University of Lübeck, 23538 Lübeck, Germany

*

Author to whom correspondence should be addressed.

†

These authors contributed equally to this work.

Nutrients 2023, 15(8), 1852; https://doi.org/10.3390/nu15081852

Submission received: 27 February 2023 / Revised: 4 April 2023 / Accepted: 7 April 2023 / Published: 12 April 2023

(This article belongs to the Section Nutrition and Metabolism)

Downloadkeyboard_arrow_down 

Browse Figures

Review Reports Versions Notes

Abstract

(1) Background: Spermidine is a biogenic polyamine that plays a crucial role in mammalian metabolism. As spermidine levels decline with age, spermidine supplementation is suggested to prevent or delay age-related diseases. However, valid pharmacokinetic data regarding spermidine remains lacking. Therefore, for the first time, the present study investigated the pharmacokinetics of oral spermidine supplementation. (2) Methods: This study was designed as a randomized, placebo-controlled, triple-blinded, two-armed crossover trial with two 5-day intervention phases separated by a washout phase of 9 days. In 12 healthy volunteers, 15 mg/d of spermidine was administered orally, and blood and saliva samples were taken. Spermidine, spermine, and putrescine were quantified by liquid chromatography–mass spectrometry (LC–MS/MS). The plasma metabolome was investigated using nuclear magnetic resonance (NMR) metabolomics. (3) Results: Compared with a placebo, spermidine supplementation significantly increased spermine levels in the plasma, but it did not affect spermidine or putrescine levels. No effect on salivary polyamine concentrations was observed. (4) Conclusions: This study’s results suggest that dietary spermidine is presystemically converted into spermine, which then enters systemic circulation. Presumably, the in vitro and clinical effects of spermidine are at least in part attributable to its metabolite, spermine. It is rather unlikely that spermidine supplements with doses <15 mg/d exert any short-term effects.

(1) 배경:

스퍼미딘은 포유류 대사에서

중요한 역할을 하는 생물학적 폴리아민입니다.

스퍼미딘 수치는 연령이 증가함에 따라 감소하며,

스퍼미딘 보충이 노화 관련 질환의 예방 또는 지연을 위해 제안되어 왔습니다.

그러나

스퍼미딘에 대한 유효한 약동학 데이터는 여전히 부족합니다.

따라서

본 연구에서는 구강 투여 스퍼미딘 보충제의 약동학을 최초로 조사했습니다.

(2) 방법:

본 연구는 무작위 배정, 위약 대조, 삼중 맹검, 두 군 교차 설계로 진행되었으며,

9일간의 세척 기간을 두고 두 개의 5일 간 개입 단계로 구성되었습니다.

12명의 건강한 자원자에게

15 mg/d의 스페르미딘을 경구 투여했으며,

혈액 및 타액 샘플을 채취했습니다.

스퍼미딘, 스퍼민, 푸트레신은

액체 크로마토그래피-질량 분석법(LC-MS/MS)으로 정량화되었습니다.

혈장 대사체는

핵자기공명(NMR) 대사체학으로 조사되었습니다.

(3) 결과:

위약 대비 스퍼미딘 보충은

혈장 내 스퍼민 수치를 유의미하게 증가시켰지만,

스퍼미딘이나 푸트레신 수치에는 영향을 미치지 않았습니다.

타액 내 폴리아민 농도에 대한 영향은

관찰되지 않았습니다.

(4) 결론:

본 연구 결과는

식이성 스페르미딘이 체내에서 스퍼민으로 사전 시스템적으로 전환된 후

체내 순환계로 들어간다는 것을 시사합니다.

체외 및 임상 효과는 적어도 일부는

그 대사산물인 스퍼민에 기인할 가능성이 있습니다.

15 mg/d 미만의 용량으로 투여된

스페르미딘 보충제는 단기적 효과를 발휘할 가능성이 낮습니다.

Keywords: 

spermidine; spermine; putrescine; autophagy; SARS-CoV-2; COVID-19

1. Introduction

Spermidine is a biogenic polyamine that plays a crucial role in cellular homeostasis, cell growth, proliferation, and autophagy [1,2,3]. Research has found a relevant role for biogenic polyamines in both health and disease [4,5]. The physiological spermidine pool is fed by cellular de novo synthesis, intestinal microbiota biosynthesis with subsequent absorption, and dietary sources [6]. In mammalian cells, spermidine is generated from its precursor putrescine or through the degradation of spermine [7], and different mechanisms of transmembrane transport contribute to the regulation of intracellular spermidine concentrations [8].

Polyamines (including spermidine) in the gastrointestinal lumen originate from food, intestinal microbiota, pancreatic-biliary secretions, and degraded intestinal cells [9,10], and dietary polyamines are proposed to be the main source of luminal polyamines [8].

The average daily nutritional intake of spermidine in European countries varies from approximately 10 to 15 mg [11]. Examples of foods with very high spermidine concentrations are cheddar cheese (200 mg/kg) and mushrooms (90 mg/kg) [12]. The dietary sources of individual spermidine intake vary widely due to individual food preferences and regional dietary habits [13,14]. The main dietary sources of spermidine are potatoes (6.4% of total), vegetable sprouts (7.3%), salad (9.8%), pears and apples (13.3%), and whole grains (13.4%) [13].

To date, little is known about the mechanisms of intestinal absorption, presystemic metabolism, and the kinetics of polyamine transport systems in humans [8]. The intestinal absorption of polyamines is complex due to different polyamine transporters [7]. After polyamine-containing foods are ingested, the majority of luminal polyamines disappear from the intestinal lumen through either passive diffusion or intraluminal degradation [15,16]. Although dietary polyamines are found in millimolar concentrations in the duodenum and jejunum after a high-dose intake, the plasma concentration reaches a maximum of 10–20 μM. Furthermore, 120 min after a polyamine-rich meal, polyamine concentrations in the intestinal lumen return to preprandial baseline levels. Thus, only low concentrations are reached in the blood (μM) despite high concentrations in food (mM) [17]. In rodent models, after oral intake, radioactively labeled dietary polyamines have been reported to reach all organs, the skeletal muscles, and even tumor tissues [18,19].

1. 소개

스퍼미딘은 세포의 항상성, 세포 성장, 증식 및 자가포식에서 중요한 역할을 하는 생체성 폴리아민입니다 [1,2,3]. 연구 결과, 생체 내 폴리아민은 건강과 질병 모두에서 중요한 역할을 하는 것으로 밝혀졌습니다 [4,5]. 생리적 스퍼미딘 풀은 세포 내 신규 합성, 장 미생물군집의 생합성 및 흡수, 식이 소스에서 공급됩니다 [6]. 포유류 세포에서 스퍼미딘은 전구체인 푸트레신에서 생성되거나 스퍼민 분해 과정을 통해 생성되며 [7], 세포 내 스퍼미딘 농도 조절에는 다양한 막 투과 메커니즘이 기여합니다 [8].

위장관 내강에 존재하는 폴리아민(스페르미딘 포함)은 식품, 장내 미생물, 췌장-담도 분비물, 분해된 장 세포에서 유래합니다 [9,10], 그리고 식이 폴리아민이 내강 폴리아민의 주요 원천으로 제안되었습니다 [8].

유럽 국가의 평균 일일 영양 섭취량은 약 10~15mg입니다[11]. 스페르미딘 농도가 매우 높은 식품의 예로는 체다 치즈(200mg/kg)와 버섯(90mg/kg)이 있습니다[12]. 개인별 스페르미딘 섭취의 식이 원천은 개인의 식품 선호도와 지역별 식습관에 따라 크게 다릅니다[13,14]. 스페르미딘의 주요 식이 원천은 감자(총량의 6.4%), 채소 싹(7.3%), 샐러드(9.8%), 배와 사과(13.3%), 전곡류(13.4%)입니다 [13].

현재까지 인간에서의 장 흡수 메커니즘, 전신 흡수 전 대사, 폴리아민 운반 시스템의 동력학에 대한 연구는 제한적입니다 [8]. 폴리아민의 장 흡수는 다양한 폴리아민 운반체로 인해 복잡합니다 [7]. 폴리아민을 함유한 식품이 섭취된 후, 장 내강에 존재하는 대부분의 폴리아민은 수동 확산이나 내강 내 분해로 인해 장 내강에서 사라집니다 [15,16]. 고용량 섭취 후 십이지장과 공장에서 폴리아민은 밀리몰 농도로 검출되지만, 혈장 농도는 최대 10–20 μM에 달합니다. 또한 폴리아민이 풍부한 식사 후 120분 후에는 장 내강 내 폴리아민 농도가 식전 기준 수준으로 돌아갑니다. 따라서 식품 내 고농도(mM)에도 불구하고 혈중 농도는 낮은 수준(μM)에 머무릅니다 [17]. 쥐 모델에서 경구 섭취 후 방사성 표지된 식이 폴리아민이 모든 장기, 골격근, 심지어 종양 조직까지 도달했다는 보고가 있습니다 [18,19].

Luminal polyamines are (at least in part) actively imported into intestinal epithelial cells (IECs) through the polyamine transport system (PTS) [20]. In IECs, the polyamines may be converted by polyamine metabolic enzymes or emitted across the basolateral membrane to the systemic circulation [15]. Moreover, paracellular passive absorption may also contribute to the uptake of luminal polyamines. Additionally, antizymes and antizyme inhibitors regulate cellular polyamine uptake [8]. However, the role of these proteins in intestinal polyamine absorption remains rather speculative [21].

Spermidine has an autophagy-inducing effect that makes it highly interesting for preclinical research [5,22]. As mechanistic studies have demonstrated, spermidine induces autophagy and mitophagy through various mechanisms. The predominant mechanisms are the inhibition of the lysine acetyltransferase EP300 and the induction of a reactive oxygen species (ROS) burst, which leads to the activation of the Ataxia-telangiectasia mutated (ATM) protein [6].

Most prominent translational research with spermidine has focused on its possible efficacy in the context of brain health/aging, oncology, and SARS-CoV-2 infections.

Regarding brain health and aging, spermidine-induced autophagy has been postulated to exert protective effects on neurodegenerative disorders and subjective cognitive decline and also to slow aging processes even in humans [23,24,25]. However, the role of spermidine in brain aging and age-related cognitive decline is complex. On the one hand, supplementation with spermidine may support healthy aging, while on the other hand, elevated spermidine plasma levels are associated with advanced brain aging, and they might serve as a potential early biomarker for Alzheimer’s disease and vascular brain pathology [26].

장 내강 폴리아민은 (적어도 일부는) 폴리아민 운반 시스템(PTS)을 통해 장 상피 세포(IECs)로 능동적으로 흡수됩니다 [20]. IECs에서 폴리아민은 폴리아민 대사 효소에 의해 전환되거나 기저측막을 통해 체내 순환계로 방출될 수 있습니다 [15]. 또한, 세포간 수동 흡수 또한 장 내강 폴리아민의 흡수에도 기여할 수 있습니다. 또한, 항자임 및 항자임 억제제는 세포의 폴리아민 흡수를 조절합니다 [8]. 그러나, 이러한 단백질이 장에서 폴리아민 흡수에 미치는 역할은 아직 추측의 단계에 머물러 있습니다 [21].

스퍼미딘은 자가포식 유도 효과가 있어 전임상 연구에서 큰 관심을 받고 있습니다 [5,22]. 기계적 연구에 따르면, 스퍼미딘은 다양한 메커니즘을 통해 자가포식 및 미토파기를 유도합니다. 주요 메커니즘은 라이신 아세틸트랜스퍼레이즈 EP300의 억제와 활성 산소 종(ROS)의 급증으로, 이는 아타시아-텔랑게이케시아 돌연변이(ATM) 단백질의 활성화를 유발합니다 [6].

스퍼미딘에 대한 가장 주목받는 번역 연구는 뇌 건강/노화, 종양학, SARS-CoV-2 감염 맥락에서의 잠재적 효능에 초점을 맞추고 있습니다.

뇌 건강 및 노화와 관련하여, 스페르미딘에 의해 유도된 자가포식은 신경 퇴행성 질환 및 주관적 인지 기능 저하에 대한 보호 효과를 발휘하고, 인간에서도 노화 과정을 늦추는 것으로 추정되고 있습니다 [23,24,25]. 그러나, 스페르미딘이 뇌 노화 및 노화 관련 인지 기능 저하에 미치는 역할은 복잡합니다. 한 편으로는 스페르미딘 보충이 건강한 노화를 지원할 수 있지만, 다른 한편으로는 혈장 내 스페르미딘 수치 상승이 뇌 노화의 진행과 연관되어 있으며, 알츠하이머 병 및 혈관성 뇌 병리의 잠재적 초기 바이오마커로 작용할 수 있습니다 [26].

In fact, the administration of exogenous spermidine extends the lifespan in various models of aging through epigenetic modifications, autophagy induction, and necrosis suppression [27]. In animal models, spermidine supplementation has had beneficial effects on brain and cognitive health [5]. However, clinical data failed to confirm‎ any positive effects of spermidine on age-related cognitive decline in contrast to preclinical studies, implying a respective effect [28]. The same also applies to the putative role of spermidine in other age-related diseases [8,29], cardiovascular health [30,31,32], nonalcoholic steatohepatitis (NASH) [33], Alzheimer’s disease [34], immune functions [35], and ophthalmology [36].

In oncology, due to its direct link with oncogenes, polyamine metabolism has long been a target for potential cancer therapeutic agents [37]. For example, recent data indicate that polyamines could play a major role in regulating the antitumor immune response [7]. Furthermore, in vivo data suggest that the efficacy of antineoplastic chemotherapies might be increased by 48-h fasting periods prior to drug application [23]. The hypothetical mechanism is fasting-mediated autophagy induction. If respective effects can be confirmed in clinical trials, autophagy-inducing compounds such as spermidine might be suitable for use in patients with contraindications against fasting. To date, however, the role of autophagy in cancer remains controversial [38].

Recently, in vitro data revealed that an infection with SARS-CoV-2 leads to the viral inhibition of spermidine synthase; as a result, the intracellular spermidine concentration decreases, and autophagy as a cellular defense mechanism is reduced [38]. Incubation of SARS-CoV-2-infected Vero cells with 333 µM spermidine reduced viral replication by 85%; furthermore, spermidine preincubation prior to SARS-CoV-2 infection reduced viral replication by 70% [38]. Based on these data, it has been hypothesized that spermidine supplementation might have a clinically relevant antiviral effect on a SARS-CoV-2 infection [39]. This hypothesis was endorsed by more current experimental results, which suggested that spermidine might play a central role in COVID-19 pathogenesis [40].

To probe these preclinical data in future clinical trials, detailed knowledge of spermidine pharmacokinetics is essential. Since SARS-CoV-2 replicates predominantly in the pharynx during the initial phase of infection, saliva concentrations of spermidine (as well as spermine and putrescine) after oral intake of spermidine are of major interest. Therefore, in addition to the blood plasma concentrations, we also analyzed the respective saliva concentrations. To the best of our knowledge, the concentrations of spermidine, spermine, and putrescine in saliva after oral intake of spermidine have never been investigated before.

Thus, to foster translational research with spermidine, we determined the pharmacokinetics of orally administered spermidine, its precursor putrescine, and its metabolite spermine, as well as the effect of its multiple-day administration on the steady-state concentrations in blood and saliva in healthy volunteers.

Additionally, also for the first time, we investigated the effect of orally administered spermidine on the human plasma metabolome using nuclear magnetic resonance (NMR) spectroscopy because spermidine has been postulated to act as a calorie restriction mimetic (CRM) in humans, which would imply effects on the plasma concentrations of various metabolites [41,42].

사실, 외인성 스퍼미딘을 투여하면 후성 유전적 변형, 자가포식 유도 및 괴사 억제를 통해 다양한 노화 모델에서 수명이 연장됩니다 [27]. 동물 모델에서 스퍼미딘 보충제는 뇌 및 인지 건강에 유익한 효과가 있는 것으로 나타났습니다 [5]. 그러나, 전임상 연구와 달리 임상 데이터는 스퍼미딘이 연령 관련 인지 기능 저하에 대한 긍정적인 효과를 확인하지 못하여, 그 효과가 불확실함을 시사합니다 [28]. 이는 스페르미딘의 다른 노화 관련 질환[8,29], 심혈관 건강[30,31,32], 비알코올성 지방간염(NASH)[33], 알츠하이머 병[34], 면역 기능[35], 안과[36]에서의 잠재적 역할에도 적용됩니다.

종양학 분야에서 폴리아민 대사는 종양 유전자와의 직접적인 연관성으로 인해 잠재적 암 치료제로 오랫동안 연구 대상이 되어 왔습니다[37]. 예를 들어, 최근 데이터는 폴리아민이 항종양 면역 반응 조절에 주요 역할을 할 수 있음을 시사합니다[7]. 또한, 체내 데이터는 약물 투여 전 48시간 금식이 항종양 화학요법의 효과를 증가시킬 수 있음을 제안합니다[23]. 가설적인 메커니즘은 단식을 통한 자가포식 유도입니다. 임상 시험에서 각 효과가 확인될 경우, 스페르미딘과 같은 자가포식을 유도하는 화합물은 단식에 금기증이 있는 환자에게 사용하기에 적합할 수 있습니다. 그러나 현재까지 암에서 자가포식의 역할은 여전히 논란이 되고 있습니다 [38].

최근 체외 데이터에 따르면 SARS-CoV-2 감염은 스페르미딘 합성 효소를 바이러스로 억제하여 세포 내 스페르미딘 농도를 감소시키고, 세포 방어 메커니즘인 자가포식을 감소시키는 것으로 밝혀졌습니다 [38]. SARS-CoV-2에 감염된 Vero 세포를 333 µM 스페르미딘과 함께 배양했을 때 바이러스 복제가 85% 감소했으며, SARS-CoV-2 감염 전 스페르미딘 사전 처리 시 바이러스 복제가 70% 감소했습니다 [38]. 이러한 데이터에 기반해 스페르미딘 보충이 SARS-CoV-2 감염에 대한 임상적으로 의미 있는 항바이러스 효과를 가질 수 있다는 가설이 제기되었습니다 [39]. 이 가설은 최근 실험 결과에 의해 지지되었으며, 스페르미딘이 COVID-19 병리학에서 중심적인 역할을 할 수 있음을 시사했습니다 [40].

미래 임상 시험에서 이러한 전임상 데이터를 탐구하기 위해

스페르미딘의 약동학에 대한

상세한 지식은 필수적입니다.

SARS-CoV-2는 감염 초기 단계에서

주로 인두에서 복제되기 때문에,

스페르미딘 경구 투여 후 타액 내 스페르미딘(스퍼민 및 푸트레신 포함) 농도는

주요 관심사입니다.

따라서

혈장 농도 외에도 해당 타액 농도를 분석했습니다.

우리의 지식 범위 내에서,

스페르미딘 경구 투여 후 타액 내 스페르미딘, 스페르민, 푸트레신 농도는

이전에 조사된 적이 없습니다.

따라서

스페르미딘을 활용한 번역 연구를 촉진하기 위해,

우리는 경구 투여된 스페르미딘,

그 전구체인 푸트레신, 및 대사산물인 스퍼민의 약동학을 결정했으며,

건강한 자원자에서 다일 투여가 혈액 및 타액의 안정 상태 농도에

미치는 영향을 조사했습니다.

또한, 스페르미딘이 인간에서 칼로리 제한 모방체(CRM)로 작용한다는 가설에 따라,

스페르미딘이 인간 혈장 대사체에 미치는 영향을 핵자기공명(NMR) 분광법을 사용하여

처음으로 조사했습니다.

이는 다양한 대사체의 혈장 농도에 영향을 미칠 수 있음을

시사합니다 [41,42].

2. Materials and Methods2.1. Study Design

The study was designed as a randomized, placebo-controlled, triple-blinded, two-armed crossover trial with a washout phase of 9 days between the two intervention phases (Figure 1). The first intervention phase took place from Monday to Friday, followed by the washout phase (the weekend and following week), and then the second intervention phase started on the following Monday.

Figure 1. Study design.

The subjects were recruited through the University of Luebeck’s email distribution list. Both supplementation and blood/saliva sampling were performed at the Institute of Nutritional Medicine. Randomization, blinding, and the production of the placebo and verum preparations were provided by the Center of Drug Absorption and Transport (C_DAT) of the University of Greifswald, Germany. Randomization was performed using random numbers generated in Microsoft Excel. With the command = random number(), a random number between 0 and 1 was generated for each subject. Subsequently, a comparison was made in blocks of two to determine which subject had the higher random number. Those with the higher random number started with the verum arm, while subjects with the lower random number started with the placebo arm. This procedure ensured an equal distribution (balance) of the two possible orders (V-P or P-V) among the 12 subjects. Thus, 6 subjects each start with verum or placebo. The capsules taken (verum or placebo) were provided by staff in Greifswald on the basis of the generated sequence for weeks 1 and 2, specifically for the subject-specific random number. Externally, verum and placebo are indistinguishable due to color-matched filling powder and capsules of the same size and color. For more information about products and capsules, see Section 2.3.

The containers were assigned on the basis of pseudonym and subject number as well as treatment week. Thus, at the study site in Lübeck, the performing staff as well as the subjects were blinded to the assignment and only received the respective prepared capsule container for week 1 or week 2. Until the analyses were completed, the samples and results were only assigned to week 1 or week 2 of the subjects. Only after the completion of the measurements was the unblinding carried out using the randomization list. Thus, analytical staff, clinical personnel, and subjects were blinded to the randomization order.

To detect possible dietary changes in the course of the study that might have influenced the results, the participants documented their meals throughout the study period in a validated nutritional log. Throughout the study course, subjects were advised to refrain from extraordinary physical work, dietary changes, and alcohol consumption of more than two standard glasses per day. In addition, the participants were advised to follow their usual activity and dietary habits.

As this study was designed as an exploratory pharmacokinetic trial, case number calculation was not possible due to a lack of data. A sample size of 12 was considered appropriate as it is established in phase 1 clinical trials. Moreover, the robustness of the data was improved by the cross-over design with placebo control.

The study was registered in the German Clinical Trials Register (DRKS) (DRKS00029397).

2.2. Study Cohort and Ethics

Study participants were recruited between December 2021 and April 2022. The inclusion criteria were age 18–45 and sufficient knowledge of the German language to understand the study instructions. The exclusion criteria were BMI < 18 or >29.9 kg/m2, acute diseases, a history of food allergy, intake of any medication, and dietary supplements during the 48-h period before the start of the study until the end of the study (except oral contraceptives), use of illegal substances, pregnancy and breast-feeding, shift work, and adherence to a vegan diet.

In total, 85 subjects were assessed for eligibility (Figure 2). After applying the inclusion and exclusion criteria and clarifying the study details, 12 subjects (eight females, four males) agreed to participate and were randomized accordingly. Informed consent was obtained from all subjects. All participants completed the study schedule. The median age of the female participants was 24 years (21–45 years), while that of the male participants was 26 years (22–28 years).

Figure 2. A modified flow chart of the randomization phases according to CONSORT 2010 [43].

2.3. Spermidine and Placebo Preparation

Commercially available dietary supplements of the brand MoleQlar™ (Rottach-Egern, Germany) based on Chlorella algae powder and soybean extract were purchased, each containing 1.5 mg of spermidine per capsule. The capsules were vegan and consisted of hydroxypropyl methylcellulose (HPMC). The content indicated on the label was verified by an independent laboratory. Five administered hard capsules corresponded to a single dose of 7.5 mg of spermidine. Similar HPMC capsules (Vcaps Plus, Lonza Capsules & Health Ingredients, Bornem, Belgium) were used for the placebo and filled with inert and similar colored capsule filler consisting of microcrystalline cellulose (MCC Pharmacell® 102, DFE Pharma, Goch, Germany) and iron oxide (Ferrum oxydatum flavum, Caesar & Loretz, Hilden, Germany).

2.4. Intervention

The intervention started with a standard dinner the evening before the first measurement day (day 0 and day 4; Figure 3).

Figure 3. Detailed sequences of the two study phases. Depending on the randomization, the placebo was used first in intervention phase 1, followed by verum supplementation in phase 2, or vice versa. Five capsules: intake of five capsules of placebo or verum with the first bite of the meal; samples: blood and saliva samples were taken; breakfast: standard breakfast; breakfast at home: self-selected, non-standardized breakfast at home; lunch: standard lunch; snack: standard snack; dinner: standard dinner; dinner at home: self-selected, non-standardized dinner at home; 100 mL of water: intake of 100 mL of water immediately after taking the sample at the respective time. After lunch, water intake was provided ad libitum. Colored days/times mark study-related activities at the study center. Days/times with study-related activities at home or without any study-related activities are not colored.

The next morning (measurement day; day 1 and day 5), placebo or verum capsules were administered to the subjects. The target dose in the verum arm of the study was 15 mg of spermidine per day (corresponding to five capsules in the morning and five in the evening). The same number of capsules were administered in the placebo arm. Subsequently, during the following 8 h, blood samples (9 mL of whole blood per sample) and saliva samples (after each blood sample) were taken under standardized dietary conditions. For the next 3 days, the respective capsules (verum or placebo) were self-administered by the participants every morning and evening, with one preprandial sampling in the morning. The second measurement day was on day 5 (similar to day 1). After a wash-out phase of 9 days, the second intervention phase started similarly to the first intervention phase.

The water intake during the measurement days was standardized until noon. The participants drank 100 mL of non-carbonated water every full hour, directly after blood and saliva samples were taken. After lunch, they had water ad libitum.

The calculation of pharmacokinetics was based on all blood and saliva samples. By contrast, only six samples of blood per subject were analyzed for the NMR metabolomics measurements (day 1 before breakfast, day 1 after snack, and day 5 after snack in both intervention phases; corresponding to samples 1, 11, and 25 or 0 h, 8 h, and 104 h, respectively).

2.5. Standard Meals

The use of standard meals ensured that the study participants consumed a comparable amount of polyamines with their food before and during the sampling on the measurement days. The standard meals were taken at the Institute for Nutritional Medicine and were composed as follows:

•  
•  
•  
•  

The meals between the measurement days were not standardized. However, study participants were instructed to refrain from relevant dietary changes during the study course. Compliance with this instruction was ensured by reviewing the nutritional logs provided by the participants.

2.6. Blood and Saliva Sampling

To ensure the validity of the time-dependent concentration determinations, blood and saliva samples were taken with a maximum time deviation of 5% of the planned time intervals. On days 1, 5, 15, and 19, blood samples were taken using a peripheral venous indwelling catheter placed in the crook of the subject’s elbow. At the end of each study day, the catheter was removed. On the shorter measurement days (days 2, 3, 4, 16, 17, and 18), blood samples were taken using a disposable cannula (butterfly). At any time point scheduled, 9 mL of peripheral blood was taken in an EDTA K3 Monovette® (Sarstedt AG & Co. KG, Nuembrecht, Germany), corresponding to a total of 50 blood samples per subject by the end of the study.

After collection, blood samples were centrifuged at 4 °C and 3000× g for 10 min. It was ensured that the samples were never at room temperature for more than 10 min prior to refrigerated centrifugation. Immediately after centrifugation, the plasma supernatant was pipetted into ice-cooled microtubes. For each sample, 1000 μL of plasma was transferred into three different microtubes and frozen at −80 °C. Two samples were used for the time-dependent concentration determination and NMR metabolomics, respectively, and one sample was used as a retained sample.

The saliva samples were placed in appropriate microtubes by the study participants immediately after the respective blood sampling and frozen at −80 °C.

2.7. Pharmacokinetics

Putrescine(1,4-diaminobutan) dihydrochloride and spermidine trihydrochloride were obtained from Acros Organics (Geel, Belgium), and spermine tetrahydrochloride was purchased from Alfa Aesar (Haverhill, MA, USA). Putrescine-d4(1,4-diaminobutan-2,2,3,3-d4) dihydrochloride, spermidine-(butyl-d8) trihydrochloride, N-(9-Fluorenylmethoxycarbonyloxy)succinimide (Fmoc-OSu), boric acid, and formic acid were obtained from Sigma-Aldrich (St. Louis, MO, USA). Acetonitrile (super gradient grade) was obtained from VWR (Radnor, PA, USA), while sodium hydroxide was provided by AppliChem (Darmstadt, Germany). All solvents (acetonitrile, water) and additives (formic acid) used for LC–MS/MS analysis were of LC–MS grade and obtained from VWR (Radnor, USA).

The determination of the concentrations of spermidine, spermine, and putrescine in plasma and saliva was performed by protein precipitation following derivatization with Fmoc-OSu based on the method of Xiong and Zhai [44]. Isotopically labeled spermidine (spermidine-d8 for spermidine and spermine) and putrescine (putrescine-d4 for putrescine) were used as internal standards.

Stock solutions, quality control spiking solutions, and internal standard solutions were prepared in methanol-water (20:80, v/v) and diluted with the same solvent unless stated otherwise. The quantitation range was 2–500 ng/mL for all polyamines in plasma. For saliva, a quantitation range of 10–2500 ng/mL was chosen for spermidine and spermine, while it was 0.4–100 µg/mL for putrescine. Each concentration of putrescine-d4 and spermidine-d8 in the internal standard solution was 2.5 µg/mL for plasma and 100 µg/mL and 2.5 µg/mL, respectively, for saliva analysis.

A detailed overview of the sample preparation can be found in Supplementary Figure S1. Briefly, the frozen (−80 °C) samples were thawed at room temperature and vortexed (IKA® VORTEX 2, IKA®-Werke GmbH & Co. KG, Staufen, Germany). The saliva samples were centrifuged for 10 min at 16,060× g (Biofuge® pico, Heraeus, Hanau, Germany), while the plasma samples were directly processed further. An aliquot of 100 µL of plasma or supernatant of centrifuged saliva, respectively, was transferred into a 1.5 mL microcentrifuge tube (Thermo Fisher Scientific, Waltham, MA, USA) and spiked with 25 µL of internal standard solution. Proteins were precipitated by adding acetonitrile containing 0.2% formic acid. After centrifugation for 10 min at 16,060× g and the transfer of an aliquot of the supernatant into a 1.5 mL amber glass vial (VWR, Radnor, PA, USA), 200 mM borate buffer with a pH of 9.0 and 5 mM Fmoc-OSu solution in acetonitrile were added. After 30 min of incubation at room temperature, the derivatization reaction was stopped by adding formic acid (100%). This extract was analyzed using a LC–MS/MS system (Shimadzu Corporation, Kyoto, Japan). The LC–MS/MS system consisted of an LC–40B X3 solvent delivery module, an SIL-40C X3 auto sampler, a CTO-40S column oven, an SPD-40 UV-Vis detector, an FCV-20AH2 valve unit, and an LC–MS-8060 mass spectrometer equipped with an ESI ionization unit.

The LC system was equipped with a Phenomenex Kinetex® 2.6 µm PS C18 100 Å 150 × 2.1 mm column (Phenomenex, Torrance, CA, USA) protected by a SecurityGuardTM ULTRA Cartridge (Phenomenex, Torrance, CA, USA), which was connected to a SecurityGuardTM ULTRA Holder (Phenomenex, Torrance, CA, USA). Acetonitrile + 0.1% HCOOH (A) and water + 0.1% HCOOH (B) were used as eluents for the mobile phase at a flow rate of 0.4 mL/min. A linear gradient was applied for the separation of the derivatized polyamines (Supplementary Table S1). The oven temperature was set to 40 °C. The total run time of the method was 12 min.

The retention times of spermidine, spermine, and putrescine were 5.60, 6.90, and 3.65 min, respectively. To prevent contamination of the ionization unit, the eluent was initially directed to a waste bottle. For the analysis of plasma samples, after 3.4 min, the valve unit changed the direction of the flow to the mass spectrometer, and after 7.4 min, the valve switched back to its initial position. For the determination of putrescine in saliva, the valve switched the flow to the MS after 3.4 min and back after 4.6 min, while for the measurement of spermidine and spermine in saliva, the eluent was directed to the MS from 5.1 to 7.4 min.

The detection of analytes was performed in the positive multiple reaction monitoring (MRM) mode (Supplementary Table S2). Furthermore, the following parameters were applied: nebulizing gas flow 3 L/min, heating gas flow 10 L/min, drying gas flow 3 L/min, interface temperature 300 °C, desolvation temperature 526 °C, desolvation line temperature 250 °C, and heat block temperature 400 °C.

Data acquisition and analysis were performed using LabSolutions (Version 5.97 SP1). The calibration curves consisted of eight calibration standards and were constructed by plotting peak area ratios of the analytes and internal standards against the concentration of the analytes. A 1/c2 weighted least squares linear regression was used for all polyamines.

The method was validated with respect to intra- and inter-day accuracy and precision at three concentration levels (QC-L, QC-M, and QC-H) in six replicates on the same day and 18 replicates over 2 days, respectively. Stability (freeze-thaw stability, long-term stability [6 weeks at −80 °C], and short-term stability [3 h at room temperature]) was eval‎uated at low (QC-L) and high (QC-H) quality control concentrations in four replicates. Furthermore, reinjection stability was investigated at three concentration levels (QC-L, QC-M, and QC-H) in six replicates after storage for 24 h in the autosampler. Plasma and saliva quality control standards were prepared by spiking QC working solutions (comprising 5% of the total volume) into the corresponding matrix. As no analyte-free plasma or saliva were available, the parameters of validation were eval‎uated by subtracting the concentration of nonspiked samples from spiked matrix samples. The matrix effect was eval‎uated by spiking the plasma and saliva of six individuals at low (QC-L) and high (QC-H) quality control concentrations and calculating the coefficient of variation (CV) of the measured concentration. As presented in Supplementary Tables S3 and S4, the method proved to be accurate and precise with respect to the aforementioned validation parameters.

Moreover, the stability investigation (data not presented) proved that the unprocessed plasma samples were stable for two freeze-thaw cycles, while unprocessed saliva samples were stable for three cycles. Unprocessed plasma and saliva samples were stable for a period of 6 weeks at −80 °C and for 3 h at room temperature. The reinjection of plasma and saliva samples after storage for 24 h in the autosampler revealed that the final extract was stable. The CV of the measured concentration of the six spiked individual plasma and saliva samples was below 15% at all investigated concentration levels (QC-L and QC-H).

2.7. 약동학

푸트레신(1,4-디아미노부탄) 디하이드로클로라이드와

스페르미딘 트리하이드로클로라이드는

Acros Organics (Geel, Belgium)에서 공급받았으며,

스페르민 테트라하이드로클로라이드는

Alfa Aesar (Haverhill, MA, USA)에서 구매했습니다.

푸트레신-d4(1,4-디아미노부탄-2,2,3,3-d4) 디히드로클로라이드,

스페르미딘-(부틸-d8) 트리히드로클로라이드,

N-(9-플루오렌일메톡시카르보닐옥시)수신이미드(Fmoc-OSu),

보릭 산, 및 포름산은 Sigma-Aldrich(St. Louis, MO, USA)에서 구입했습니다.

아세토니트릴(슈퍼 그라디언트 등급)은 VWR (Radnor, PA, USA)에서,

나트륨 하이드록사이드(NaOH)는 AppliChem (Darmstadt, Germany)에서 공급받았습니다.

LC–MS/MS 분석에 사용된 모든 용매(아세토니트릴, 물) 및 첨가제(포름산)는

LC–MS 등급으로 VWR (Radnor, USA)에서 구입했습니다.

혈장 및 타액 중 스페르미딘, 스페르민, 푸트레신의 농도는

Xiong과 Zhai [44]의 방법을 기반으로

Fmoc-OSu를 이용한 유도체화 후 단백질 침전법을 통해 측정되었습니다.

동위 원소 표지된 스페르미딘(스페르미딘-d8 및 스페르민-d8)과 푸트레신(푸트레신-d4)은

내부 표준으로 사용되었습니다.

표준 용액, 품질 관리 첨가 용액 및 내부 표준 용액은 메탄올-물 (20:80, 부피비율)로 준비되었으며,

달리 명시되지 않은 경우 동일한 용매로 희석되었습니다.

플라즈마 내 모든 폴리아민의 정량 범위는

2–500 ng/mL였습니다.

타액의 경우 스페르미딘과 스페르민은

10–2500 ng/mL, 푸트레신은 0.4–100 µg/mL로 정량 범위가 설정되었습니다.

내부 표준 용액의 푸트레신-d4 및 스퍼미딘-d8 농도는

혈장 분석 시 각각 2.5 µg/mL, 타액 분석 시 각각 100 µg/mL 및 2.5 µg/mL로 설정되었습니다.

시료 준비의 상세한 개요는 보조 그림 S1에 수록되어 있습니다. 간단히 설명하면, 동결된 (−80 °C) 시료는 실온에서 해동된 후 IKA® VORTEX 2 (IKA®-Werke GmbH & Co. KG, Staufen, Germany)로 혼합되었습니다. 타액 시료는 Biofuge® pico (Heraeus, Hanau, Germany)에서 16,060× g로 10분간 원심분리되었으며, 혈장 시료는 직접 추가 처리되었습니다. 원심분리된 혈장 또는 타액 상청액의 100 µL를 각각 1.5 mL 마이크로 원심분리 튜브(Thermo Fisher Scientific, Waltham, MA, USA)에 옮기고 내부 표준 용액 25 µL를 첨가했습니다.

단백질은 0.2% 포름산이 포함된 아세토니트릴을 추가하여 침전시켰습니다. 16,060× g에서 10분간 원심분리 후 상층액을 1.5 mL 암갈색 유리 바이알(VWR, Radnor, PA, USA)로 옮기고, pH 9.0의 200 mM 보레이트 완충액과 아세토니트릴에 용해된 5 mM Fmoc-OSu 용액을 추가했습니다. 실온에서 30분간 반응시킨 후, 포름산(100%)을 추가하여 유도체화 반응을 중단합니다. 이 추출물은 LC–MS/MS 시스템(Shimadzu Corporation, Kyoto, Japan)을 사용하여 분석되었습니다. LC–MS/MS 시스템은 LC–40B X3 용매 공급 모듈, SIL-40C X3 자동 샘플러, CTO-40S 컬럼 오븐, SPD-40 UV-Vis 검출기, FCV-20AH2 밸브 유닛, 및 ESI 이온화 유닛이 장착된 LC–MS-8060 질량 분석기로 구성되었습니다.

LC 시스템에는 Phenomenex Kinetex® 2.6 µm PS C18 100 Å 150 × 2.1 mm 컬럼(Phenomenex, Torrance, CA, USA) SecurityGuardTM ULTRA 카트리지(Phenomenex, Torrance, CA, USA)로 보호되었으며, 이는 SecurityGuardTM ULTRA 홀더(Phenomenex, Torrance, CA, USA)에 연결되었습니다. 이동상 용매로 아세토니트릴 + 0.1% HCOOH (A)와 물 + 0.1% HCOOH (B)를 사용했으며, 유속은 0.4 mL/min으로 설정되었습니다. 유도체화된 폴리아민(보충 표 S1)의 분리를 위해 선형 그라디언트를 적용했습니다. 오븐 온도는 40 °C로 설정되었습니다. 방법의 총 분석 시간은 12분입니다.

스페르미딘, 스페르민, 및 푸트레신인의 유지 시간은

각각 5.60, 6.90, 및 3.65분이었다.

이온화 장치의 오염을 방지하기 위해 용출액은 처음에 폐기용 병으로 유도되었습니다.

혈장 시료 분석을 위해 3.4분 후 밸브 유닛이 유동 방향을 질량 분석기로 전환했으며,

7.4분 후 밸브는 초기 위치로 돌아갔습니다.

타액 내 푸트레신의 농도를 측정하기 위해 밸브는

3.4분 후 MS로 유체 흐름을 전환하고

4.6분 후 다시 전환했습니다.

반면

타액 내 스퍼미딘과 스퍼민을 측정하기 위해 용출액은

5.1분부터 7.4분까지 MS로 직접 유입되었습니다.

분석물의 검출은 양이온 다중 반응 모니터링(MRM) 모드에서 수행되었습니다(보충 표 S2). 또한 다음과 같은 파라미터가 적용되었습니다: 분무 가스 유량 3 L/min, 가열 가스 유량 10 L/min, 건조 가스 유량 3 L/min, 인터페이스 온도 300 °C, 탈용매화 온도 526 °C, 탈용매화 라인 온도 250 °C, 열 블록 온도 400 °C.

데이터 수집 및 분석은 LabSolutions(버전 5.97 SP1)를 사용하여 수행되었습니다. 교정 곡선은 8개의 교정 표준으로 구성되었으며, 분석물과 내부 표준의 피크 면적 비율을 분석물 농도에 대해 그래프화하여 작성되었습니다. 모든 폴리아민에 대해 1/c² 가중 최소 제곱 선형 회귀 분석이 적용되었습니다.

이 방법은 3개 농도 수준(QC-L, QC-M, QC-H)에서 동일한 날에 6회 반복 및 2일 동안 18회 반복으로 intra- 및 inter-day 정확도와 정밀도를 평가하여 검증되었습니다. 안정성(동결-해동 안정성, 장기 안정성[−80°C에서 6주], 단기 안정성[실온에서 3시간])은 저농도(QC-L) 및 고농도(QC-H) 품질 관리 농도에서 4회 반복 측정으로 평가되었습니다. 또한, 자동 샘플러에서 24시간 보관 후 3개 농도 수준(QC-L, QC-M, QC-H)에서 6회 반복하여 재주입 안정성을 조사했습니다. 혈장 및 타액 품질 관리 표준은 총 부피의 5%를 차지하는 QC 작업 용액을 해당 매트릭스에 첨가하여 준비했습니다. 분석물 없는 혈장 또는 타액이 없었기 때문에, 검증 매개변수는 첨가되지 않은 시료의 농도를 첨가된 매트릭스 시료의 농도에서 빼서 평가되었습니다. 매트릭스 효과는 6명의 개인의 혈장과 타액을 저농도(QC-L) 및 고농도(QC-H) 품질 관리 농도로 첨가하고 측정된 농도의 변동계수(CV)를 계산하여 평가되었습니다. 보충 표 S3 및 S4에 제시된 바와 같이, 해당 방법은 앞서 언급된 검증 매개변수에 대해 정확하고 정밀한 것으로 입증되었습니다.

또한 안정성 조사(데이터 미제시) 결과, 처리되지 않은 혈장 샘플은 2회 동결-해동 사이클 동안 안정적이었으며, 처리되지 않은 타액 샘플은 3회 사이클 동안 안정적이었음을 확인했습니다. 미처리 혈장 및 타액 샘플은 −80°C에서 6주간, 실온에서 3시간 동안 안정적이었습니다. 자동 샘플러에서 24시간 보관 후 재주입한 혈장 및 타액 샘플의 최종 추출물은 안정적이었습니다. 6명의 개인 혈장 및 타액 샘플에 첨가된 농도의 측정 농도 CV는 모든 조사된 농도 수준(QC-L 및 QC-H)에서 15% 미만이었습니다.

2.8. NMR Metabolomics

Plasma samples were analyzed for their metabolites and lipoproteins by NMR spectroscopy according to Bruker’s standardized and certified in vitro diagnostic research protocol (IVDr; Bruker BioSpin, Billerica, MA, USA). Samples were processed and measured following a strict standard operating procedure, which was previously described by Dona et al. [45]. In brief, thawed aliquots were homogenized with 75 mM sodium phosphate buffer (1/1, v/v, pH 7.4) by manual shaking and transferred into a 5 mm NMR tube. Prior to analysis, samples were stored at 6 °C in an automated SampleJet™. NMR analysis was performed on a Bruker 600 MHz Avance III HD nuclear magnetic resonance spectrometer with a TXI probe at 37 °C (310 K). Temperature calibration and quantification, as well as water suppression performance, were checked, and settings were optimized before starting the experiments if necessary. Based on the IVDr protocol, a one-dimensional (1D) NOESY experiment (pulse program: noesygppr1d) and a 1D Carr–Purcell–Meiboom–Gill spin-echo experiment (CPMG; pulse program: cpmgpr1d) for the suppression of macromolecules and proteins were measured per sample. The eval‎uation and quantification of selected parameters were performed according to Bruker quantification in plasma/serum B.I.Quant-PS 2.0.0 and Bruker IVDr Lipoprotein Subclass Analysis (B.I.-LISA). In total, 39 metabolites (+two technical additives) and 112 lipoprotein parameters were quantified automatically (Bruker BioSpin).

2.8. NMR 대사체학

혈장 샘플은 Bruker의 표준화 및 인증된 체외 진단 연구 프로토콜(IVDr; Bruker BioSpin, Billerica, MA, USA)에 따라 NMR 분광법으로 대사체 및 지단백질을 분석했습니다. 시료는 Dona 등 [45]에서 이전에 설명된 엄격한 표준 운영 절차에 따라 처리 및 측정되었습니다. 간략히 설명하면, 해동된 시료는 75 mM 나트륨 인산염 완충액(1/1, v/v, pH 7.4)과 수동 흔들기로 균일화한 후 5 mm NMR 튜브에 옮겨졌습니다. 분석 전 시료는 자동화된 SampleJet™에서 6 °C에서 보관되었습니다. NMR 분석은 Bruker 600 MHz Avance III HD 핵자기공명 분광계와 TXI 프로브를 사용하여 37°C (310 K)에서 수행되었습니다. 온도 교정 및 정량화, 물 억제 성능을 확인했으며, 필요 시 실험 시작 전에 설정을 최적화했습니다. IVDr 프로토콜에 따라 각 시료당 거대분자 및 단백질 억제를 위한 1차원(1D) NOESY 실험(펄스 프로그램: noesygppr1d)과 1D Carr–Purcell–Meiboom–Gill 스핀 에코 실험(CPMG; 펄스 프로그램: cpmgpr1d)을 측정했습니다. 선택된 매개변수의 평가 및 정량은 Bruker 정량화 방법(플라즈마/혈청 B.I.Quant-PS 2.0.0) 및 Bruker IVDr 리포프로틴 하위 클래스 분석(B.I.-LISA)에 따라 수행되었습니다. 총 39개의 대사체(+2개의 기술적 첨가물)와 112개의 리포프로틴 매개변수가 자동으로 정량화되었습니다(Bruker BioSpin).

2.9. Data Analysis and Statistics

The data analysis was conducted using GraphPad version 9.4.0. The normal distribution of the pharmacokinetic data was tested using the D’Agostino–Pearson (omnibus K2) and Anderson–Darling methods. The area under the curve (AUC) from the first to the last sample of the respective intervention phase (AUC0-tlast), the maximal achieved concentration (cmax), and the time at which it appeared (tmax) were examined to compare the verum and placebo interventions. The data were plotted for each determined substance in plasma and saliva with their mean and SD for each sampling time point.

The individual profiles of spermidine and spermine in plasma can be found in Supplementary Figures S2–S7. The correlation between the plasma and saliva values of each substance was examined using Spearman’s rank correlation coefficient. Regarding the fasting values, a verum value was compared with the corresponding placebo value, depending on the normal distribution, using a paired t test or a Wilcoxon matched-pairs signed-rank test. A Friedmann test was performed to classify the fasting verum values after multiple administrations of spermidine. The same was performed for the fasting placebo values to compare the verum values with the placebo trend. The differences were considered significant at p ≤ 0.05.

Sample 11 of the verum arm was compared with sample 11 of the placebo arm (8 h after first administration), and sample 25 of the verum arm was compared with sample 25 of the placebo arm (the last samples of each intervention phase). The respective samples from the two intervention phases (t = 0 h) were also compared to test the setup and baseline determinations. In addition, a Wilcoxon matched-pairs signed-rank test was performed to detect significant differences between the two intervention arms. All metabolites were plotted together in a forest plot to provide an overview of the changes in sample 11 of the verum arm versus sample 11 of the placebo arm, as well as in sample 25 of the verum arm versus sample 25 of the placebo arm (Figure 4 and Figure 5). The relative error with the placebo sample as a reference value was used. Outlier values were detected using the Grubbs test (ESD method) and removed from the analysis; thus, a total of seven data points were removed. Furthermore, a linear regression was applied to detect possible correlations between the change in AUC0-tlast spermine plasma and metabolome. Therefore, the verum–placebo difference for both AUC0-tlast and metabolic parameters from the last sample of the respective intervention phase was calculated (sample 25).

Figure 4. Concentrations (mean ± SD, n = 12) of spermidine (left) and spermine (right) in plasma with time point t = 0 h corresponding to first intake of respective allocated treatment and respective baseline concentration obtained immediately before intake (A) at first measurement day of verum or placebo treatment, respectively (which can be day 1 or day 15 due to randomization), 0–8 h after first administration of verum or placebo; (B) first samples of days 1, 5, 15, and 19 and days 2, 3, 4, 16, 17, and 18 (all fasted matutinal samples of each intervention phase); (C) day 5 and day 19 (last measurement day of each intervention phase which can be verum or placebo due to randomization), 96–104 h after multiple administration of verum or placebo.

Figure 5. Spermine concentration in plasma measured for 0–104 h was determined by AUC0-tlast after placebo and verum interventions.

3. Results3.1. Samples

Due to unforeseen events (dizziness and personal appointments), a total of four plasma and two saliva samples could not be taken. Only five (0.833%) of a total of 600 sampling times (50 per subject) were outside the targeted time interval. The largest time deviation was 7 min, while the shortest deviation was 2 min. Thus, more than 99% of the samples were taken within the scheduled time interval. No special relevance was attributed to these five samples, and they were analyzed like the other samples. One saliva sample and one plasma sample were not eval‎uable due to the analytical failure of the LC–MS/MS system.

3.2. Pharmacokinetics in Blood Plasma

Initially, the mean concentrations of spermidine, spermine, and putrescine in the plasma and saliva in the placebo intervention of all 12 subjects combined over all time samples were determined as follows: spermidine (plasma) 12.41 ng/mL (SD 1.869), spermine (plasma) 5.102 ng/mL (SD 0.4646), and putrescine (plasma) 7.403 ng/mL (SD 2.130); spermidine (saliva) 236.5 ng/mL (SD 181.5), spermine (saliva) 201.0 ng/mL (SD 76.02), and putrescine (saliva) 4.343 µg/mL. Subsequently, concentrations of spermidine, spermine, and putrescine in plasma and saliva for both the placebo and verum interventions were compared for AUC0-tlast, cmax, and tmax. Comparing placebo and verum, no significant differences between the plasma concentrations of spermidine and putrescine were found (Figure 4, Table 1, Figure S8, Table S7). For spermine, plasma AUC0-tlast was significantly increased in the verum group (p = 0.0282).

Table 1. The plasma spermidine and spermine mean and SD of AUC0-tlast, cmax, tmax, and p value for placebo versus verum.

Interestingly, four of the 12 subjects still had higher AUC0-tlast plasma concentrations for spermine in the placebo intervention compared to the verum intervention (data not presented). Noteworthy, fasting plasma spermine concentrations were higher than the postprandial values in both the placebo and verum arms. The mean values of the fasting samples and the remaining postprandial samples of each subject were averaged to a mean value across the 12 subjects. The mean of the fasted matutinal sample’s placebo intervention for spermine in plasma was 5.720 ng/mL (SD 1.205); for the fasted matutinal samples of the verum intervention, it was 6.502 ng/mL (SD 1.492); for the postprandial samples of the placebo intervention, it was 4.940 ng/mL (SD 1.124); and for the postprandial samples of the verum intervention, it was 4.965 ng/mL (SD 1.324). Investigating whether oral spermidine supplementation increased fasting plasma spermine levels (fasted matutinal concentrations of verum vs. placebo at the respective time point; Wilcoxon signed rank test or paired t test depending on normal distribution), no significant difference was found (Table 2); however, a trend of a rising difference was observable over the study duration.

Table 2. p values from the Wilcoxon signed rank test or paired t test derived from the comparison of the respective fasted matutinal placebo measurements with the associated fasted matutinal verum measurements of spermine in plasma.

For both study arms, a Friedman test (nonparametric ANOVA) was performed to compare the fasted matutinal spermine plasma levels at 0, 24, 48, 72, and 96 h with each other within the specific study treatment (verum or placebo). Within the placebo treatment, a p value of 0.7847 (ns) was found, indicating no difference between matutinal concentrations on the days during the placebo treatment. A Dunn’s multiple comparison test did not find a significant change throughout the days when concentrations at 24, 48, 72, and 96 h were compared with starting spermine concentrations at 0 h during placebo treatment. Within the verum group, a p value of 0.0823 (ns) was found, indicating a strong trend of a change in fasted matutinal spermine concentrations during spermidine intake without reaching statistical significance. A Dunn’s multiple comparison test also found no significant change throughout the days when concentrations at 24, 48, 72, and 96 h were compared with starting spermine concentrations at 0 h, even though an increase in matutinal concentrations was observed during verum treatment.

3.3. Salivary Pharmacokinetics

An analysis of salivary concentrations found no significant changes in AUC0-tlast, cmax, and tmax (Figure S9, Table S8, Figure 6, and Table 3) for spermidine, spermine, or putrescine.

Figure 6. Salivary concentrations (mean ± SD, n = 12) of spermidine (left) and spermine (right); (A) day 1 and day 15 (first measurement day of each intervention phase), 0–8 h after first administration of verum or placebo; (B) first sample days 1, 5, 15, and 19 and days 2, 3, 4, 16, 17, and 18 (all fasted matutinal samples of each intervention phase); (C) day 5 and day 19 (last measurement day of each intervention phase), 96–104 h after multiple administration of verum or placebo.

Table 3. Salivary spermidine and spermine AUC0-tlast, cmax, tmax, and p value for placebo versus verum. Data are presented as the mean (SD).

The correlations between plasma and saliva concentrations of spermidine, spermine, and putrescine were significantly poor (Spearman r < 0.5, p > 0.05; (Figure 7 and Figure 8)).

Figure 7. Correlation of saliva and plasma concentrations of spermidine (A) and spermine (B) for the verum intervention over all time points. Data were analyzed using Spearman’s rank correlation coefficient.

Figure 8. Correlation of saliva and plasma concentrations of spermidine (A) and spermine (B) for the placebo intervention over all time points; data were analyzed using Spearman’s rank correlation coefficient.

3.4. NMR Metabolomics

When comparing metabolite concentrations in the blood plasma, no significant differences were observed between the verum and placebo interventions (Figure S12). This also applied in particular to the concentrations of glucose and acetone, both of which are metabolites that have been postulated to change due to the application of calorie restriction mimetics such as spermidine.

As demonstrated above, spermine AUC0-tlast plasma concentrations of the mean values significantly differed when comparing the verum and placebo groups. Since four of the 12 subjects exhibited a spermine AUC0-tlast in the placebo intervention greater than in the verum intervention, a correlation analysis of the metabolome and spermine AUC0-tlast was performed. In the linear regression, the differences between the verum and placebo interventions in AUC0-tlast values of spermine plasma concentrations for each subject were compared with the corresponding differences in metabolite concentrations of the sample 25 (the last value of the corresponding intervention phase). A significant difference was found for LDL-3/4/5, LDL FC, LDL PL, LDL-3/4/5 Chol, LDL-4 FC, LDL-3/4/5 PL, LDL-3/4/5 ApoB, and HDL-1 ApoA2.

4. Discussion

4.1. Brief Synopsis of Key Findings

To the best of our knowledge, this is the first randomized placebo-controlled pharmacokinetic study with spermidine in humans. Previous data on the pharmacokinetics of spermidine are either based on animal studies only or on clinical trials that do not meet the standards of pharmacokinetic studies. Most strikingly, we found that oral intake of spermidine at 15 mg/d for 5 days increased spermine levels in the plasma but not levels of spermidine or putrescine. Similarly, a nonsignificant trend indicated that spermidine supplementation increased fasted matutinal spermine plasma concentrations but not respective spermidine and putrescine concentrations.

4.2. In Vivo and Ex Vivo Data

Available pharmacokinetic data for spermidine are mainly limited to in vivo and animal models and suggest contradictory results. In a rodent ex vivo model, the portal vein recovery of 13C-labeled spermidine after jejunal instillation was 61–76% within 20 min [19]. In rats, intragastrically administered 14C-labeled polyamines were found to be rapidly absorbed in the small intestine, passed into systemic circulation, and distributed to different tissues [9]. In a different setting with rats, gavage of a liquid spermidine/spermine mixture followed by whole blood sampling from both the portal vein and vena cava inferior led to significantly increased concentrations of both spermidine and spermine in whole blood samples, peaking at 1 h after oral administration [46].

In mouse models, prolonged administration of polyamine-rich chow was found to (partly sex-dependently) increase both spermidine and spermine blood levels [17,47]. In contrast, Soda et al. reported that polyamine-rich chow in mice significantly increased spermine blood concentrations but not spermidine concentrations, which is in line with our results [48].

Interestingly, early in vivo experiments in rats with dietary polyamines from the 1990s revealed differential effects of presystemic eliminations. After intragastrical administration, only 11–15% of putrescine was recovered in intestinal tissue, but 87–96% of spermidine and 79–82% of spermine were recovered [49]. Based on these data, it has been suggested that most dietary putrescine is degraded by diamine oxidase and polyamine oxidase, which are both abundant in intestinal tissue. However, dietary spermidine and spermine are almost quantitatively absorbed without catalytic breakdown. Further presystemic processes after absorption remain unclear.

4.3. Pharmacokinetic Data from Clinical Trials

Previous analyses of spermidine pharmacokinetics in humans are scarce; published data are not based on randomized placebo-controlled interventions but rather on noninterventional data or food-based interventions. In 2018, the results of a randomized, placebo-controlled, double-blind phase II trial with supplementation of a polyamine-rich plant extract were reported [47]. For 3 months, 30 participants received a supplement providing a daily dose of spermidine (1.2 mg), spermine (0.6 mg), and putrescine (0.2 mg). Blood polyamine concentrations did not differ between placebo controls and spermidine-supplemented individuals. This result can easily be explained by the underdosing of the supplement.

More recently, confirm‎ing data from rodent experiments [48], in a nonrandomized intervention trial, Soda et al. investigated the effect of a polyamine-rich food (natto) on spermidine and spermine blood concentrations of young male subjects [50]. In this trial, natto consumption was found to increase the mean spermine levels in whole blood, while blood spermidine levels remained unchanged despite increased dietary intake. Our results are in line with these findings, despite being based on a completely different methodology. Thus, taking together respective data from rodents [48], human food-based data [50], and our results presented here, one can postulate that spermidine, after oral intake, rather acts as a prodrug that is converted to spermine by spermine synthase. Consequently, previously described health-related effects after spermidine supplementation are likely to be mediated not by spermidine itself but by increased spermine concentrations. Notwithstanding, it might be possible that systemically distributed spermine is subsequently reconverted to spermidine by spermine oxidase (SMOX) or spermine synthase (SMS) in peripheral tissues.

4.4. Salivary Concentrations of Polyamines

To our knowledge, no data on the salivary pharmacokinetics of spermidine, spermine, and putrescine after oral supplementation of spermidine have been published before. In our study, neither the salivary AUC0-tlast, cmax, nor tmax differed significantly between the verum and placebo treatments for the three polyamines investigated. Furthermore, no significant correlation between plasma and salivary concentrations of spermidine, spermine, or putrescine was detected.

Polyamines, including spermidine and putrescine, reportedly have diverse functions in the bacterial physiology of the oral microbiota [51]. However, the role of the oral microbiota in systemic polyamine metabolism has yet to be elucidated [52]. To date, it remains speculative whether a differential oral microbiota composition may contribute to the differential effects of oral spermidine intake on salivary polyamine concentrations.

As described earlier, a reduction of the viral replication of SARS-CoV-2 by 85% was demonstrated in vitro at a concentration of 333 µM spermidine, encouraging discussions about the preventive or even therapeutic potential of spermidine in the context of COVID-19 [38]. Our results, however, do not support this approach, as even the highest individual salivary spermidine concentration of 3191 ng/mL, corresponding to 22.0 µM, is smaller by more than a factor of 10 than the aforementioned effective in vitro concentration. Thus, oral spermidine supplementation does not seem to be an effective approach for reducing the risk of an infection with SARS-CoV-2. Higher spermidine concentrations in the pharynx could possibly be achieved using alternative application methods, such as a spray, mouth rinsing and gargling, or inhalation.

Furthermore, limited data are available on salivary concentrations of spermidine, spermine, and putrescine without oral spermidine supplementation. Cooke et al. investigated the time profiles of putrescine, cadaverine, indole, and skatole upon waking and during the day [53]. In their study, the highest concentrations of putrescine (mean 33.0 µg/mL) were measured immediately upon waking. In the present study, the highest concentrations of polyamines were also observed in fasted matutinal salivary probes. After breakfast and brushing, the salivary putrescine concentration fell to 7.0 µg/mL, while during the day an increase was observed. These high matutinal salivary polyamine concentrations are probably caused by bacterial putrefaction [53]. Proteins and peptides present in the oral cavity are broken down into amino acids, which are further decomposed by various enzymes. For example, ornithine can be converted to putrescine by ornithine decarboxylase. High amounts of putrescine may be processed to spermidine by spermidine synthase and further to spermine by spermine synthase [54]. During sleep, spermidine, spermine, and putrescine could accumulate in the oral cavity, and, after waking up, eating meals and oral cleaning would lead to a decrease in salivary polyamine concentrations.

Li et al. measured the salivary concentration of various polyamines and amino acids in five healthy volunteers before and after they brushed their teeth [55]. Before brushing, the concentration of spermidine in saliva was 3.40–10.84 µmol/L (493.9–1575 ng/mL), whereas after brushing it was in a range of 1.61–6.41 µmol/L (233.9–931.1 ng/mL). Salivary putrescine concentrations were in the range of 0.54–133.36 µmol/L (47.6–11,756 ng/mL) before and 0.26–3.64 µmol/L (22.9–320.9 ng/mL) after brushing teeth. Several research groups have investigated the salivary concentrations of polyamines as a possible biomarker for various types of cancer [56,57,58,59]. Usually, cancer patients have been demonstrated to have elevated salivary polyamine concentrations. However, even in healthy subjects, the variability of salivary polyamine concentrations across different studies is apparently rather high. Venza et al. investigated the salivary polyamine concentrations in 50 healthy subjects and divided the subjects into five different groups according to age and/or gender [60]. No association between age or gender and salivary polyamine concentrations was found. The determined average salivary concentrations of the groups of 107–110 ng/mL for spermidine and 118–121 ng/mL for spermine were in the same order of magnitude as those in our study. However, Venza et al. found average salivary putrescine concentrations of 24–29 ng/mL, much lower concentrations than our study found. In another study, the mean salivary concentrations of spermidine and spermine of 14 healthy subjects were below 10 ng/mL, while the mean salivary putrescine concentrations were approximately 130 ng/mL with noticeable interindividual variability [56]. Takayama et al. found the mean salivary concentrations of 61 healthy subjects to be approximately 0.55 nmol/mL (79.9 ng/mL) for spermidine, 0.055 nmol/mL (11.1 ng/mL) for spermine, and 15.5 nmol/mL (1366 ng/mL) for putrescine [57]. Dame et al. reported average salivary concentrations of spermidine of 3.21 µmol/L (466.3 ng/mL), of spermine of 0.63 µmol/L (127.5 ng/mL), and of putrescine of 136.76 µmol/L (12,055 ng/mL) for 16 healthy volunteers [61]. Overall, both the inter- and intraindividual variability of salivary polyamine concentrations was quite high in our study, but pharmacokinetics related to supplementation were not eval‎uable.

4.5. Relevance of Presystemic Conversion of Spermidine

Although it has been speculated that intestinal polyamine uptake might be similar in rodents and humans, our data indicate that—different from rodent data—dietary spermidine is not quantitatively recovered in the systemic circulation but is subject to extensive presystemic metabolism. These findings are supported by previous studies. For instance, Soda et al. demonstrated millimolar concentrations of dietary polyamines in the duodenum and jejunum in humans after high-dose oral intake yielded maximal plasma concentrations of only 10–20 μM [17]. Most likely, after enterocytic uptake, spermidine is almost completely converted to spermine, either in the enterocytes or by hepatic first-pass metabolism. The mechanisms of spermidine degradation to spermine were investigated and described in detail elsewhere [8]. Predominantly, spermidine is converted to spermine by spermine synthase. Alternatively, spermidine is converted to acetyl-spermidine by spermidine/spermine-N(1)-acetyltransferase, which is then degraded to putrescine by N(1)-acetyl-spermine/spermidine oxidase or microsomal acetylpolyamine oxidase. Thus, putrescine is both a biological precursor of spermidine and its metabolite. Nonetheless, we observed no changes in putrescine concentrations in saliva or plasma.

Both the enterocytic and hepatic conversion of spermidine to spermine might explain the increased spermine concentration in blood plasma after oral spermidine intake, while spermidine concentrations remained unchanged. Alternative explanations could be a complete intraluminal degradation, a lack of intestinal absorption, or a rapid distribution to tissues and organs with peripheral uptake rates being similar to intestinal absorption, which would leave blood concentrations at a constant level. Since none of these mechanisms can explain the observed increase in spermine levels over time, our data strongly suggest that in humans, dietary spermidine is at least partly presystemically converted into spermine, which then occurs in the systemic circulation. Since no spermine pharmacokinetics were eval‎uable, these processes are probably concealed by other metabolic or distributing processes, as discussed above.

4.6. Role of the Intestinal Microbiota

Evidently, the pharmacokinetic relationship between oral spermidine intake and spermidine/spermine plasma concentrations is not simply additive, and the underlying mechanisms remain rather speculative. Obviously, dietary spermidine is subject to extensive presystemic metabolism. In this context, it has been suggested that the intestinal microbiota is involved in the presystemic metabolism of dietary polyamines. Matsumoto et al. reported that the administration of probiotic bifidobacteria increased spermine—but not spermidine—concentrations in the feces of animals and humans [62,63]. In fact, polyamine production by intestinal bacteria is a main source of exogenous polyamines, and dietary arginine (a precursor of polyamine biosynthesis) was demonstrated to increase both colonic putrescine concentration and blood levels of spermidine and spermine [63]. However, as dietary polyamines have been reported to be quantitatively absorbed in the duodenum and the proximal jejunum, the involvement of the colonic microbiota in the presystemic metabolism of dietary spermidine appears to be rather unlikely.

4.7. NMR Metabolomics

As demonstrated above, a positive correlation of plasma spermine concentrations with several cholesterol fractions was found. A causal effect due to spermidine supplementation cannot simply be postulated, since no significant differences were observed between the verum and placebo interventions (Figure S12). While these results could be the basis for future research, they should not be discussed further in the context of the present pharmacokinetic study.

4.8. Clinical and Research Implications

Since spermidine supplementation has been postulated as a possible means of therapeutically targeting several diseases [3,7], translational approaches with spermidine are a hot topic. Therefore, our results have considerable significance for future clinical research.

First, our data demonstrate that a daily dose of 15 mg of spermidine significantly increases blood concentrations of spermine. In previous trials, daily spermidine doses of 0.9–1.2 mg have been used without any change in blood concentrations of polyamines or any clinical efficacy [47,64]. Taking our data into account, those results may have been a consequence of simple underdosing. It is rather unlikely that spermidine supplements with daily doses below 15 mg exert any effect, since even the effect observed in this study was rather small. Thus, in future trials, daily doses of at least 15 mg should be used. While it is not impossible that long-term supplementation of low-dose spermidine (<15 mg/d) exerts any effect through epigenetic modulations or metabolic cascades, such effects remain highly speculative.

Second, it can be postulated that the in vitro and clinical effects of spermidine are not solely due to spermidine itself but may also be attributed to its metabolite, spermine, which is formed in the body prior to or in peripheral tissues [64]. The concept of spermidine mainly representing the prodrug of biologically active spermine is further supported by the pathophysiology of Snyder–Robinson syndrome (SRS) [65]: SRS is a very rare X-linked recessive condition caused by a mutation in the spermine synthase gene located at Xp.21.3-p22.12. This mutation causes incorrect splicing that results in an inactive truncated protein, leading to the loss of most spermine synthase activity (>90%). Affected patients exhibit numerous adverse health effects, including intellectual disability, muscle hypoplasia, facial dysmorphism, renal abnormalities, and seizures [66,67]. Most interestingly, spermine is significantly reduced in SRS, whereas spermidine is substantially increased.

Third, the overall view of preclinical and clinical data, including the results presented here, suggests that the mechanisms of intestinal absorption as well as the mechanisms of presystemic spermidine metabolism are different in rodents and humans, thereby directly affecting the biological effects. Therefore, special caution is required when transferring experimental animal data with spermidine to humans. On the other hand, spermine might represent the predominant form of transport of spermidine in the blood, with dietary spermidine being absorbed into enterocytes and intracellularly converted to spermine, which is then extruded into the systemic circulation through the basolateral membrane, followed by reconversion to spermidine in tissues and organs. It is well known that the conversion of spermidine to spermine is not a one-way street but rather possible in both directions [8]. Nonetheless, these processes were not visible in spermine pharmacokinetics after spermidine supplementation.

Fourth, oral spermidine intake of 15 mg/d does not affect salivary polyamine concentrations. Preclinical data implying an oropharyngeal effect of spermidine may not be translated into recommendations to supplement spermidine, as effective doses have yet to be determined. Systemic doses of 15 mg/d do not seem to be appropriate for yielding pharmacological concentrations in the saliva.

Fifth, our findings challenge the concept of epidemiological spermidine studies based on statistical correlations between the spermidine content of food and biological effects without considering the variability of presystemic spermidine metabolism or dietary spermine intake.

4.9. Limitations of the Study

There are a number of limitations that should be considered in interpreting the current findings. Since this study was designed as an exploratory pharmacokinetic trial, case number calculation was not possible due to a lack of data. Accordingly, a sample size of 12 was considered appropriate as it was established in phase 1 clinical trials. Therefore, it cannot be excluded that the study was not sufficiently powered to detect minor concentration changes. However, the robustness of the data was improved by the cross-over design with placebo control, which can be considered the baseline.

Validated analytic results of spermidine, spermine, and putrescine can be estimated as robust but were performed in plasma after protein precipitation. It is unclear whether the substances could be transported intracellularly in blood cells or bound to plasma proteins in relevant amounts, which would have led to an underestimation of the complete amount present in whole blood. It is also possible that there might be some significant effects of spermidine supplementation if taken by special subgroups, in higher doses, or simultaneously with certain foods. Moreover, the intervention phases were rather short (5-day interventions), and it cannot be excluded that a long-term intake of spermidine would show deviating results. The latter seems unlikely, as Soda et al. obtained very similar results on spermine concentrations with similar amounts of supplemented spermidine, even after long-term administration [50].

5. Conclusions

The results presented here for the first time provide a pharmacokinetic basis for future translational research on spermidine. Oral spermidine intake of 15 mg/d for 5 days significantly increased spermine levels in the plasma but did not affect spermidine or putrescine levels. Our data strongly suggest that dietary spermidine is presystemically converted into spermine, which then occurs in the systemic circulation. Consequently, we postulate that the in vitro and clinical effects of spermidine are (at least in part) not attributable to spermidine itself but rather to its metabolite, spermine. It is rather unlikely that spermidine supplements with doses <15 mg/d exert any effect. Moreover, even spermidine doses of 15 mg/d do not affect salivary polyamine concentrations; in particular, pharmacological spermidine concentrations in the saliva that, due to preclinical data, might be effective at inhibiting oropharyngeal SARS-CoV-2 replication are not remotely achieved.

Finally, epidemiological studies that have correlated dietary spermidine intake with biological effects are challenged by our results, as such studies have not considered the variability of presystemic spermidine metabolism and dietary spermine intake.

5. 결론

본 연구에서 처음으로 제시된 결과는 

스페르미딘에 대한 미래의 전환 연구에 약동학적 기반을 제공합니다. 

5일 동안 하루 15mg의 경구 스페르미딘 섭취는 

혈장 내 스페르민 수치를 유의미하게 증가시켰지만, 

스페르미딘이나 푸트레신 수치에는 영향을 미치지 않았습니다. 

우리의 데이터는 

식이성 스페르미딘이 체내에서 스페르민으로 사전 대사되어 이후 

체내 순환계에 존재한다는 것을 강력히 시사합니다. 

따라서 

우리는 스페르미딘의 체외 및 임상 효과가 (적어도 일부는) 

스페르미딘 자체에 기인하는 것이 아니라 

그 대사산물인 스페르민에 기인한다고 추론합니다. 

15 mg/일 미만의 스페르미딘 보충제는 

어떤 효과도 발휘하지 않을 가능성이 높습니다. 

또한, 

15 mg/d의 스페르미딘 용량도 

타액 폴리아민 농도에 영향을 미치지 않습니다. 

특히, 

전임상 데이터에 따라 구인두 SARS-CoV-2 복제를 억제하는 데 효과적일 수 있는 

타액 내 약리학적 스페르미딘 농도는 전혀 달성되지 않습니다.

마지막으로, 

식이 스페르미딘 섭취와 생물학적 효과를 연관시킨 역학 연구는 

우리 연구 결과에 의해 도전받고 있습니다. 

이러한 연구들은 

식이 스페르민 섭취와 전신 흡수 전 스페르미딘 대사 변이를 

고려하지 않았기 때문입니다.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/nu15081852/s1: Figure S1: Overview of the steps for the preparation of plasma and saliva samples for LC–MS analysis. (IS—internal standard, ACN—acetonitrile, Fmoc-OSu—N-(9-Fluorenylmethoxy-carbonyloxy)succinimide). Figure S2: Individual graphs of subjects 1, 2, and 3 for plasma spermidine in both the verum and placebo interventions, divided respectively into left: 0–8 h after administration, middle: all fasted matutinal samples, and right: 96–104 h of the intervention phases. Figure S3: Individual graphs of subjects 4, 5, and 6 for plasma spermidine in both the verum and placebo interventions, divided respectively into left: 0–8 h after administration, middle: all fasted matutinal samples, and right: 96–104 h of the intervention phases. Note: different y-axis for subject 5. Figure S4: Individual graphs of subjects 7, 8, and 9 for plasma spermidine in both the verum and placebo interventions, divided respectively into left: 0–8 h after administration, middle: all fasted matutinal samples, and right: 96–104 h of the intervention phases. Note: different y-axis for subject 8. Figure S5: Individual graphs of subjects 10, 11, and 12 for plasma spermidine in both the verum and placebo interventions, divided respectively into left: 0–8 h after administration, middle: all fasted matutinal samples, and right: 96–104 h of the intervention phases. Figure S6: Individual graphs of subjects 1, 2, and 3 for plasma spermine in both the verum and placebo interventions, divided respectively into left: 0–8 h after administration, middle: all fasted matutinal samples, and right: 96–104 h of the intervention phases. Figure S7: Individual graphs of subjects 4, 5, and 6 for plasma spermine in both the verum and placebo interventions, divided respectively into left: 0–8 h after administration, middle: all fasted matutinal samples, and right: 96–104 h of the intervention phases. Note: different y-axis for subject 5. Figure S8: Individual graphs of subjects 7, 8, and 9 for plasma spermine in both the verum and placebo interventions, divided respectively into left: 0–8 h after administration, middle: all fasted matutinal samples, and right: 96–104 h of the intervention phases. Figure S9: Individual graphs of subjects 7, 8, and 9 for plasma spermine in both the verum and placebo interventions, divided respectively into left: 0–8 h after administration, middle: all fasted matutinal samples, and right: 96–104 h of the intervention phases. Figure S10: Concentrations of putrescine in plasma, mean and standard deviation of all 12 subjects combined: (A) day 1 and day 15 (first measurement day of each intervention phase), 0–8 h after first administration of verum or placebo; (B) first sample on days 1, 5, 15, and 19 and days 2, 3, 4, 16, 17, and 18 (all fasted matutinal samples of each intervention phase); (C) day 5 and day 19 (last measurement day of each intervention phase), 96–104 h after multiple administrations of verum or placebo. Figure S11: Concentrations of putrescine in saliva, mean and standard deviation of all 12 subjects combined: (A) day 1 and day 15 (first measurement day of each intervention phase), 0–8 h after first administration of verum or placebo; (B) first sample days 1, 5, 15, and 19 and days 2, 3, 4, 16, 17, and 18 (all fasted matutinal samples of each intervention phase); (C) days 5 and day 19 (last measurement day of each intervention phase), 96–104 h after multiple administrations of verum or placebo. Figure S12: Forest plot based on the relative error with a placebo sample as the reference value: (A) metabolites; (B) cholesterol; (C) triglyceride; (D) lipoprotein particles; (E) free cholesterol; (F) phospholipides; (G) apolipoprotein B; (H) apolipoprotein A1; (I) apolipoprotein A2. Table S1: Gradient conditions for the used LCMS method for the determination of polyamines. Table S2: MRM transitions used for quantification with the corresponding collision energies. Table S3: Immediate and intermediate accuracy and precision of the determination of polyamines in human plasma. Table S4: Immediate and intermediate accuracy and precision of the determination of polyamines in human saliva. Table S5: Individual AUC 0-tlast, cmax, tmax, and mean p values for spermidine plasma in both the verum and placebo interventions. Table S6: Individual AUC 0-tlast, cmax, tmax, and mean p values for spermine plasma in both verum and placebo interventions. Table S7: Plasma putrescine mean and SD AUC 0-tlast, cmax, tmax and p-value placebo versus verum. Table S8: Saliva putrescine mean and SD AUC 0-tlast, cmax, tmax, and p value in both verum and placebo interventions.

Author Contributions

S.S. and E.W. contributed equally to this paper; conceptualization: M.S., F.S., E.W., W.W. and M.G.; methodology: M.S., E.W., F.S. and M.G.; validation: S.S., P.S., F.S. and E.W.; formal analysis: F.S., E.W., S.S. and H.O.; investigation: M.S. and E.W.; resources: C.S., M.S., E.W. and W.W.; data curation: E.W., S.S. and M.G.; writing—original draft preparation: E.W. and M.S.; writing—review and editing: A.K., E.W., M.S., S.S., M.G. and P.S.; visualization: E.W.; supervision: M.S.; project administration: M.S.; and funding acquisition: M.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the University of Luebeck, Luebeck, Germany (protocol code: 20-292, approved on 13 August 2020).

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

Not applicable.

Acknowledgments

We thank Noreen Neuwirth, Heidi Schlichting, Friederike Schmidt, and Pia Witt for their excellent technical assistance.

Conflicts of Interest

The authors declare no conflict of interest.

References

1. Pegg, A.E. Functions of Polyamines in Mammals. J. Biol. Chem. 2016, 291, 14904–14912. [Google Scholar] [CrossRef] [PubMed]