Genome and clonal hematopoiesis stability contrasts with immune, cfDNA, mitochondrial, and telomere length changes during short duration spaceflight

Abstract Background The Inspiration4 (I4) mission, the first all-civilian orbital flight mission, investigated the physiological effects of short-duration spaceflight through a multi-omic approach. Despite advances, there remains much to learn about human adaptation to spaceflight's unique challenges, including microgravity, immune system perturbations, and radiation exposure. Methods To provide a detailed genetics analysis of the mission, we collected dried blood spots pre-, during, and post-flight for DNA extraction. Telomere length was measured by quantitative PCR, while whole genome and cfDNA sequencing provided insight into genomic stability and immune adaptations. A robust bioinformatic pipeline was used for data analysis, including variant calling to assess mutational burden. Result Telomere elongation occurred during spaceflight and shortened after return to Earth. Cell-free DNA analysis revealed increased immune cell signatures post-flight. No significant clonal hematopoiesis of indeterminate potential (CHIP) or whole-genome instability was observed. The long-term gene expression changes across immune cells suggested cellular adaptations to the space environment persisting months post-flight. Conclusion Our findings provide valuable insights into the physiological consequences of short-duration spaceflight, with telomere dynamics and immune cell gene expression adapting to spaceflight and persisting after return to Earth. CHIP sequencing data will serve as a reference point for studying the early development of CHIP in astronauts, an understudied phenomenon as previous studies have focused on career astronauts. This study will serve as a reference point for future commercial and non-commercial spaceflight, low Earth orbit (LEO) missions, and deep-space exploration.


Introduction
Human spaceflight frequency has increased over the last decade, and with it, our grasp of its effects on human physiology.Despite this, there is much to learn about how humans adapt to the unique challenges of the space en vironment.T he space environment includes unique stressors, such as microgravity ( e.g.weightlessness ) , immune system perturbations, and space radiation exposure.Determining the effects of spaceflight on the human genome is imper ativ e to support long-term human presence in space.Additionally, a host of physiological adaptations occur when humans tr av el into space, r anging fr om cardiov ascular and musculoskeletal deconditioning, vision changes ( e.g.spaceflight associated neuro ocular syndrome ) , immune suppression, and metabolic changes, among others [ 1 ].Mor eov er, incr easing e vidence highlights the systemic immune dysregulation crewmembers experience from spaceflight exposure.Indeed, there appears to be a reduction in T cell fr equency, suppr essed cytotoxic function, as well as fluctuations in cytokine concentrations, with incr eased expr ession of TNF-α, interleukin-8, interleukin-1r a, thr ombopoietin, VEGF, and v arious c hemokines ( CCL2, CCL4, CXCL5, etc. ) associated with spaceflight [ 2 ].
To date, almost all space missions have been led by professional astr onauts fr om gov ernment space pr ogr ams.Mor e r ecently, ho w ever, SpaceX launched a crew of four civilian astronauts on the Inspiration4 ( I4 ) mission, marking a new era for human space exploration.The present I4 mission study aims to investigate the impact of short-duration spaceflight on civilians, along with the post-flight and long-term biomedical, immunologic, and genetic alterations resulting from spaceflight.This includes r a pid telomeric r esponses, wher eby telomer e elongation occurs as a function of spaceflight, follo w ed b y shortening upon return to Earth [ 3 ].Both unusually short and long telomeres have been associated with adverse health effects involving aging and a ge-r elated pathologies, suc h as cardiov ascular disease and cancer.Furthermor e, a v ariety of lifestyle factors and envir onmental exposures influence telomere length.Previous studies have also shown changes in clonal hematopoiesis, where a small number of hematopoietic stem cells ( HSCs ) gain a clonal adv anta ge thr ough the acquisition of somatic mutations.Clonal hematopoiesis is associated with cardiovascular disease, the de v elopment of acute m yeloid leuk emia ( AML ) , and incr eased ov er all mortality [ 4 , 5 ].The NASA Twins Study shed light on telomeric responses and the c hr onic inflammatory state resulting from long-duration spaceflight and provided the first measures of telomere length and clonal hematopoiesis in astronauts [ 3 , 6 ].Mutations in epigenetic r egulators, suc h as DNMT3A and TET2, were found at increased rates in both career astronaut twins when compared to their civilian counter parts.A r etr ospectiv e study of 14 astronauts who flew shuttle missions has also shown an ele v ated pr esence of genetic abnormalities in CHIP-driver genes in the astronaut population [ 3 , 5 , 7 ].
Here we investigated immune adaptations, telomere length dynamics, cell-fr ee DNA r elease, genomic stability, single-cell tr anscriptomic anal ysis, and bioc hemical ada ptations of the I4 cr e wmembers tr av eling into lo w er earth orbit over the course of 3 da ys , to determine the effects of short-duration spaceflight.It is imper ativ e to study these adaptations in this context for future space missions, as prolonged space envir onment exposur e could exponentiall y incr ease the r ate of m utational burden [ 8 ], be it due to the inflammatory milieu r esulting fr om spaceflight or the particular challenges of the space environment, such as microgravity and radiation exposure .Here , we leveraged the I4 mission to examine the de v elopment of genetic abnormalities in the astronauts by studying their physiological response to the demands of spaceflight and quantifying the genetic changes that are associated with spaceflight.

Informed consent and ethics
This study was completed following a ppr opriate ethical guidelines according to the Declaration of Helsinki, ICHGCP, and local regulations, as applicable, from each potential subject or each subject's legally authorized re presentati ve prior to participating in the r esearc h study.The pr otocol, the ICF, other written material given to the patients, and any other r ele v ant study documentation were submitted to and a ppr ov ed by the a ppr opriate ethics committee .T his study was conducted under a protocol reviewed and a ppr ov ed by the a pplicable ethics committees and investigations were undertaken by scientifically and medically qualified persons, where the benefits of the study were in proportion to the risks.All subjects were consented at an informed consent briefing ( ICB ) at SpaceX ( Ha wthorne , C A ) , and samples were collected and processed under the approval of the Institutional Re vie w Board ( IRB ) at Weill Cornell Medicine, under Protocol 21-05023569.All cr e w members provided written informed consent for data and sample sharing.
I4 launc hed fr om K ennedy Space Center's Launc h Complex 39A and tr av eled into Low-Earth Orbit acr oss a thr ee-day mission, r eac hing an orbital altitude of a ppr oximatel y 364 miles and ultimately splashed down into the Atlantic Ocean.

DN A extr action for qPCR-based assessment of telomere length
Three 3 mm circular punches were cut from the Whatman 903 Pr otein Sav er Cards ( cat# WHA10534612 ) containing blood samples using an Integra Miltex Standard Biopsy Punch ( cat# 12-460-406 ) and placed into a 1.5 mL microcentrifuge tube with sterile tw eezers.Samples w er e pr epar ed using the Qia gen QIAamp DNA Investigator Kit ( cat# 56 504 ) following the manufacturer's isolation of total DNA from FTA and Guthrie Cards protocol.The quantification of DNA in each sample was determined through fluorometric quantification with the Qubit 4 Fluorometer ( Thermo Fisher Scientific, cat# Q33238 ) and the 1X dsDNA HS Assay Kit ( cat# Q33231 ) .DN A samples w er e sent to Color ado State Univ ersity for multiplex qPCR analysis.

Whole genome extraction and sequencing
Genomic DN A w as obtained from the cell pellet of a cell-free DNA blood collection tube ( Str ec k, cat# 230470 ) using the QI-Aamp Blood Maxi Kit ( Qiagen, cat# 51192 ) , and then shipped to Element Biosciences for library preparation.The extracted DNA was quantified using Thermo Fisher Qubit dsDNA HS Assay Kit ( cat# Q238253 ) and 8 samples wer e pr epar ed using the KAPA Hy-perPrep Kit and KAPA Unique-Dual Indexed Adapter Kit ( cat# 8 861 919 702 ) .The DNA libraries were quantified using Thermo Fisher Qubit dsDNA HS Assay Kit ( cat# Q32854 ) and sized using Agilent High Sensitivity DNA Kit ( cat# 5067-4626 ) .
The 8 DNA libraries generated with the KAPA HyperPrep Kit wer e pr ocessed using Adept Library Compatibility Kit ( Element Biosciences, Cat# 830-00003 ) , individually circularized with 0.5pmol ( 30 μL of 16.67nM ) input, and quantified using the kitprovided qPCR standard and primer mix.The libr aries wer e pooled into 4 separate 2-plex pools, each denatured and sequenced on Element AVITI system ( Element Biosciences, Part #88-00 001 ) using 2 × 150 paired end reads with indexing.Primary analysis was performed onboard the AVITI sequencing instrument.

W hole genome/cfDN A preprocessing
Blood and plasma samples were subjected to whole genome and cfDNA short read sequencing as detailed abo ve .Resultant FASTQ files wer e v alidated using FastQC ( v0.11.9 ) and MultiQC ( v1.13 ) .Read adapters were trimmed at 3 and 5 ends for low quality using Trim Galore ( v0.6.5 ) , lo w er quality reads were classified and r emov ed, r etaining onl y those r eads with length > = 25bp, and ph read quality > = 20.Reads were aligned against the hg38 human r efer ence genome with BWA MEM ( v.0.7.15 ) and subjected to standard QC and deduplication pr ocedur es as a part of Sentieon's TNscope ( v202010 ) DN Aseq w orkflo w [ 9 ].

W hole genome/cfDN A/single cell v ariant calling
Aligned and pr epr ocessed r eads wer e subjected to the TNScope variant calling pipeline.Calls were filtered using Fisher's exact test and subsetted to SNP variants using samtools ( v1.16.1 ) , and filtered by triallelic sites, short tandem repeats, read quality, and read position bias using BCFtools ( v1.16 ) .Varient-Effect-Predictor VEP ( v107 ) was utilized for annotation of variants and further filtering based on predicted impact of mutations.Resulting coordinates wer e pr ocessed into allele and gene frequency matrices, and visualized in R using the tidyverse ( v1.3.2 ) suite of packages.

cfDN A extr action and sequencing
cfDN A w as isolated from 500 μL aliquots of plasma from cfDNA blood collection tubes ( Str ec k, #230470 ) .cfDN A w as extracted fr om eac h cr e w member fr om all timepoints ( 4 cr e w members, 6 timepoints, 24 total extractions ) .cfDN A w as extracted using Qiagen's QIAamp ccf DN A/RN A Kit and eluted in 15 μL Qiagen Elution Buffer per sample.Yield was measured for each sample using Thermo Fisher Qubit 1X dsDNA HS Assay ( cat# Q33231 ) .
Entir e extr act v olume w as used as input for libr ary pr epar ation using NEBNext Ultra II DNA Library Preparation Kit for cfDNA protocol.Each sample w as bar coded using NEBNext Multiplex Oligos for Illumina ( Unique Dual Index UMI Ada ptors-96 r eactions ) .Final library was eluted in 30 μL and checked for concentration using Thermo Fisher Qubit 1X dsDNA HS Assay ( cat# Q33231 ) .Fr a gment sizes were determined using Agilent's Tapestation 2100 and D1000 r ea gents and Scr eenTa pe, with r esulting av er a ge fr a gment size ∼380 bp ( 0.25 pmol ) of each sample.
A total of 24 cfDNA libraries generated with the NEBNext Ultra II DNA Libr ary Pr epar ation kit wer e pr ocessed using Adept Libr ary Compatibility Kit ( Element Biosciences, Cat# 830-00003 ) .Each library w as cir cularized individually with an input range of 0.2-0.5 pmol ( 30 μL of 6.67-16.67nM ) based on linear library yields .T he final circularized libraries were quantified using qPCR standard and primer mix and pooled into 2 separate 4-plex pools.Each 4plex pool was denatured and sequenced on Element AVITI system ( Element Biosciences, Part #88-00001 ) using 2 × 147 paired reads with 19 bp UMI/index 1 and 8 bp index 2.Primary analysis was performed onboard the AVITI sequencing instrument.

Clonal hematopoiesis targeted variant calling
Genomic DN A w as obtained from the cell pellet of a cell-free DNA blood collection tube ( Str ec k, cat# 230 470 ) using the QIAamp Blood Maxi Kit ( Qiagen, cat# 51 192 ) .All samples from the testing and validation cohort were sequenced using a custom designed DNA sequencing assay ( DB0188, VariantPlex, ArcherDX ) .This panel ca ptur es the nine genes most commonly mutated in solid tumor patients following ther a peutic r adiation including the full exonic regions of five genes ( DNMT3A, TET2, ASXL1, TP53, CHEK2 ) and targeted exonic regions of four genes ( JAK2, SRSF2, SF3B1, PPM1D ) .Libr aries wer e pr epar ed fr om 250 ng gDNA using the VariantPlex protocol ( ArcherDx Inc., Boulder, CO, USA ) which utilizes Anc hor ed Multiplex PCR ( AMP ) tec hnology to gener ate tar get-enric hed sequencing-r eady libr aries.Follo wing DN A fr a gmentation ligation with a universal ArcherDx molecular barcode ( MBC ) adapter is performed, which tags each DNA molecule with a unique molecular index ( UMI ) and allows for unidirectional amplification of the sample using gene-specific primers .T he resulting libraries were sequenced using a NovaSeq 6000 instrument ( Ilumina ) , as per manufacturer's instructions .T he methods for variant identification are comparable to those in Novetsky et al. 2023 [ 10 ].The variant calling pipeline included UMI consensus building follo w ed b y utilizing Mutect2, VarDictJ av a, Lofr eq2, Pindel in parallel for variant identification.Normalization was performed via LeftTrimAndAlign, and a panel of normals of 27 children and young adults was used with a bonferr oni-corr ected p value to identify true putative mutations in our astronaut cohort.A pileup analysis as performed in Bolton et al. 2020 [ 11 ] was used to query variants in longitudinal follow-ups.A variant list was used to scrutinize raw BAM files and the P oN samples , a fisher's exact test and Bonferr oni-corr ected p v alue wer e used to determine a noise threshold.Variants surpassing this threshold in at least one sample were further examined in the remaining samples, and if statistically validated, the variants were aggregated to construct the final v ariant list.Variant-Effect-Pr edictor VEP ( v107 ) and SnpEff ( v4.3 ) was used for annotation of variants.Data wrangling, tidying, and visualizations were performed using R ( v4.1.2) , Rstudio ( v2021.09.2 ) and libraries ( Tidyv erse, Dpl yr, data.table,ggplot2 ) .

cfDN A fr agment anal ysis
Fr a gment size distribution was calculated using the bamPEFr a g-mentSize tool from the deepTools Python package ( v3.5.1 ) .Levels of cfDNA ( read counts ) originating from different chromosomes were normalized by c hr omosome length and total number of reads in the library generating a Read per Kilobase per Million r eads ( RPKM ) measur ement.The fr action of cell-fr ee mtDNA r elativ e to c hr omosomal cfDN A in plasma w as compared and visualized in R using the tidyverse ( v1.3.2 ) suite of packages.

cfDNA tissue of origin deconvolution
The enrichment of cfDNA fragments from various tissues was calculated by read coverage depletion analysis at transcription starting sites ( TSSs ) to estimate nucleosome positioning and infer gene expression.The pipeline is described in detail in Bezdan et al. 2020 [ 7 ].The resulting nucleosome periodicity was correlated with ( 1 ) per-tissue gene expression reference matrix retrieved from the Human Pr oteome Ma p ( HPM; Kim et al. 2014 [ 12 ] ) or ( 2 ) individual astronaut pseudo-bulk expression of different cell subpopulations extr acted fr om PBMC scRNAseq dataset.In both cases the tissue/subpopulation-periodicity correlations were ranked by the value of Pearson's correlation coefficient, clustered ( utilizing Ward method with Euclidean distances ) and visualized in R using the tidyverse ( v1.3.2 ) and ComplexHeatmap ( v2.14.0 ) packages.

Single-cell gene expression analyses
Longitudinal single cell data was processed in R using the Seur at pac ka ge ( v4.3.0 ) to normalize, scale and cluster cell populations.Cell identities were determined through computational gating parameters of inclusion based on gene expression of k e y mark ers, similar to gating from Fluorescent-Activated Cell Sorting.CD8 + T cells were selected from the PBMC population by filtering CD3D + CD8A + positive cells.CD4 + T cells from CD3D + CD8A + cells, CD14 + monocytes from CCR2 + CD14 + cells, CD16 + monocytes from CD14 + CD16 + cells, NK cells from NCAM1 + CD3-and NCR3 + CD3-cells and DCs from CD86 + , CD83 + cells.

Telomere length measurement ( MMqPCR )
We first examined DNA from dried blood spots ( DBS ) , which were collected from the I4 crew members before ( L-92, L-44, L-3 ) , during ( FD1, FD2, FD3 ) , and after ( R + 1, R + 45, R + 82 ) spaceflight, from which DN A w as isolated, and telomer e length measur ed via m ultiplexed quantitative PCR ( MMqPCR ) ( Fig. 1 A ) .Telomere elongation was observed in all four I4 cr e wmembers during spaceflight, as compared to their pre-and post-flight normalized means ( Fig. 1 A ) .Telomere length shortened rapidly upon return to Earth in 3 of the 4 astronauts ( post-flight ) , and overall continued to shorten over the course of the recovery period.Repeated-measurements ANOVA was performed on the data to assess the ov er all impact of time on telomere length.Post hoc pairwise comparisons identified pre-flight vs in-flight ( P < 0.001 ) and pre-flight vs R + 1 ( P < 0.02 ) as timepoints between which significant differences in telomere length occurred.
Intriguingly, despite the short duration of the I4 mission ( 3 days of orbital flight ) , these results correlated with findings from the NASA Twins Study ( One Year Mission astronaut ) , as well as with astronauts on ∼6 month missions onboard the International Space Station ( ISS ) [ 3 , 13 ].

Cell-free DNA
Next, cell-free DN A ( cfDN A ) from the I4 crew's plasma was extr acted fr om Str ec k Cell-fr ee DNA BCT tubes and sequenced to compar e to pr e-flight baseline, post-flight, and r ecov ery period r esponses.Extraction and pre-processing of plasma were not implemented during the flight, so all samples were collected on Earth to mitigate cfDNA contamination from apoptotic blood cells during sample transportation ( Supplememtary Fig. S1A & B, see online supplementary material ) [ 8 ].We analyzed plasma mitochondrial cfDN A ( cf-mtDN A ) r elativ e to the c hr omosomal cfDNA fr action ( Fig. 1 B & C, S upplememtary Table S3 , see online supplementary material ) as a potential biomarker for long-duration spaceflight ada ptations.In this mission, cf-mtDNA le v els did not rise significantly following the return to Earth ( R + 1 ) , possibly due to the short flight dur ation.Cf-mtDNA enric hment was onl y observ ed in the twin astronauts on the ISS after months of exposure to the space en vironment.T he marker le v els display high betweensample heterogeneity, and the increase in pre-flight measurements may be connected to the I4 cr e w's pr e-flight pr epar ation ( e.g.high-altitude training ) .
Giv en that cfDNA fr a gments encompass one nucleosome ( ∼150 bp ) , we inspected depletion in read coverage distribution around the transcription start sites ( TSS ) as a measurement of gene expr ession [ 6 , 14 ].We a pplied cov er a ge normalization specific to the genomic neighborhood and completed Fourier transformation to determine nucleosome footprints based on signal periodicity [ 7 ].The cfDNA fr a gment tissue and cell-type-of-origin enric hment of fr a gments wer e corr elated with either tissue-specific expression signatur es fr om the Human Pr oteome Ma p ( HPM ) [ 12 ] or individual astr onaut expr ession pr ofiles fr om peripher al blood [ 15 ] ( Fig. 1 D & E, S upplememtary Fig. S1C , S upplememtary Tables S1 & S2 , see online supplementary material ) .The most r epr esented sequences were of hematopoietic origin, while a small increased pr esence of cfDNA fr a gments fr om both innate and ada ptiv e immune cells was observed post-landing and during the recovery period.Inter estingl y, ther e was a significant increase in the cfDNA originating from all immune cells in all astronauts over a month following their return to Earth ( R + 82 ) , indicating a degree of heterogeneity in cell lysis over time and a possible long-term response.

Clonal hematopoiesis of indeterminate potential ( CHIP ) assessment
While CHIP has been pr e viousl y studied in astronauts, the I4 cohort provided an opportunity to study the physiological impact  S1 , see online supplementary material.( E ) Cross-examination of cfDNA origin deconvolution using each of I4 astronaut cell-subpopulation-specific expression mark ers, deri ved from peripheral blood ( PBMC ) dataset, supports the results in D. We note an incr eased pr esence of innate and ada ptiv e imm une cfDNA in R + 1 and R + 82 when compar ed to all other timepoints .T he extended data is a vailable in S upplememtary Tables S2 , see online supplementary material.
of spaceflight in a civilian cr e w that had ne v er under gone orbital flight.Indeed, the monitoring of this cr e w pr ovides us with a unique opportunity to understand the beginning of the already established relationship between career astronauts and CHIP.Of note, the NASA Twins Study assessed the mutational burden of long-duration spaceflight in CHIP genes and a retrospective study established a burden of CH-related mutations in the career astronaut population [ 3 , 4 ] and we sought to examine these same CHIP genes for the I4 cr e w .T o address the variant allele frequency ( VAF ) of CHIP genes, we performed deep targeted sequencing ( > 15,000x ) of 10 known epigenetic regulators and CHIP-associated genes and assessed the mutational burden of short duration spaceflight in the I4 astronaut cohort.Notably, the mutational burden of the CHIP-associated genes did not change as a function of spaceflight and remained stable for up to 6 months post spaceflight ( Fig. 2 A ) .No known deleterious mutations arose in the I4 cohort post-flight, nor did an y pr e-existing somatic mutations increase in allele frequency significantly ( Fig. 2 B & C ) .Putati ve dri ver mutations were found in two out of the four astronauts but their VAF remained comparable at all timepoints, indicating stability of the VAF from the spaceflight.
To give context for these data, we next collected additional samples for both subjects in the NASA Twins Study.Subject TW ( long duration spaceflight ) presented mutation TET2.p.Cys1273Tyr in a bulk PBMC sample at an increased VAF of ∼ 0.075 ( Fig. 2 D ) .Subject HR ( ground control ) presents the mutation DNMT3A.p.Trp698Ter at an increased VAF of ∼ 0.086 and the mutation DNMT3A.p.Asp856Gly not previously found at a VAF of ∼ 0.05 ( Fig. 2 E ) .Both astronauts presented increased clonality when compared to previous timepoints .T heir mutational burden was still greater than a ge-matc hed contr ols, as established in the original study.Ho w e v er, the incr eased m utational burden is an expected finding, given the association between aging and CHIP [ 16 ] and there's no evidence of them having any clinical effects to date.

Longitudinal comparison of whole genome muta tions re veals genomic stability months after short-dur a tion orbital flight
To assess the long-term consequences of spaceflight on the genomic stability of astronauts, we conducted whole-genome sequencing ( WGS ) and variant calling pre-and post-flight.Comparison of single nucleotide variant ( SNV ) and indel presence between timepoints ( L-44 and R + 45 ) for each astronaut indicated no significant changes for genes with the highest mutational burden ( Fig. 3 A ) , epigenetic regulator genes ( S upplememtary Fig. S3 , see online supplementary material ) or other gene sets of interest.The analysis of variant annotations ( Fig. 3 A & C ) revealed that most identified variants are within intronic regions and have no significant effects.We also have noted no disproportional mutational load for variants post-flight ( Fig. 3 D ) .We performed variant calling in single nucleus RNA-seq ( snRNA-seq ) libr aries and compar ed it across all timepoints and against WGS for genes with read cover a ge > 10x ( S upplememtary Fig. S3 B & C, see online supplementary material ) .Whole genome cov er a ge of the snRNA-seq libraries v aried br oadl y acr oss timepoints, but did not show any significant evidence of genomic instability within our cohort.

Spaceflight induces cell-type-specific changes in gene expression profiles
We next examined gene expression changes across immune cell types due to spaceflight.Cell type-specific genes were used to fil-ter cell types computationally.CD8 + T cells, and CD4 + T cells, CD14 + monocytes, CD16 + monocytes, dendritic cells ( DCs ) and natural killer ( NK ) cells were filtered from PBMCs at six different timepoints: L-92, L-44, L-3, R + 1, R + 45, and R + 82.

Discussion
The I4 mission is the first study to date to analyze the effect of short duration spaceflight on civilians using a compr ehensiv e m ulti-omics a ppr oac h.Our findings, including telomer e length dynamics, CHIP-related clonal expansion, WGS genomic stability, cfDNA cell lysis analysis, and immune cell longitudinal gene expr ession pr ofiling, br oadl y contextualize the physiological burdens of short duration spaceflight.This study adds key CHIP and genetic data on astr onauts, whic h ar e limited, and can serve as a r efer ence for futur e spaceflight planning, short missions around Earth's orbit, and longer-duration missions into deep space.
The NASA Twins Study was the first to report spaceflightspecific telomere elongation in humans, with one astronaut experiencing telomere elongation during his one-year mission, rapid telomere shortening upon return to Earth, and telomere length r ecov ery to near pre-flight baseline values over the following months, although many more short telomeres after spaceflight than befor e wer e also observ ed [ 3 ].Luxton et al. reported similar spaceflight-associated telomere length dynamics in 3 unrelated astronauts on ∼ 6-month missions onboard the ISS [ 13 ].It was hypothesized that such dramatic shifts in telomere length dynamics were associated with c hr onic exposur e to the space radiation envir onment and r epr esented an ada ptiv e r esponse to c hr onic oxidativ e dama ge, specificall y to telomer es, wher eby the alternativ e lengthening of telomeres ( ALT ) pathway is transiently activated in normal somatic cells [ 13 ].Cytogenetic evidence of heterogeneous  telomere lengths and DNA damage responses were also reported, and-together with our cfDNA enrichment analyses-suggested incr eased DNA dama ge and ar e consistent with space r adiation exposure and increased senescence-associated foci.Similar changes in telomere length dynamics observed in the I4 crew as a function of spaceflight indicate that telomere length alterations are fast-acting as associated with spaceflight, even for missions of short ( e .g. da ys ) dur ation.Circulating cell-fr ee DNA ( cfDNA ) fr a gments, originating fr om v arious tissues and the immune system, offer a non-inv asiv e method for assessing astronauts' dynamic imm une r esponses to spaceflight-induced physiological str ess.As an emerging biomarker, cfDNA concentration and molecular profile yield valuable insights into stress responses in unique spaceflight environments despite its heterogeneous and sensitive nature.
Our study indicates that the impact of short-duration spaceflight on cfDNA concentration in astronaut plasma is subtler than long duration missions, but increased innate and adaptive immune system activity is still apparent.Moreover, cfDNA enrichment from all immune cells persists post-flight, suggesting a dela yed response in volving immune cell turnover and active DNA r epair mec hanisms .T his phenomenon warrants further in vestigation as it is imper ativ e for both short and long duration spaceflight to elucidate the temporal changes to immune cell types in order to better predict and prevent future risks.

Effects of short and long dur a tion spaceflight on CHIP genomic alter a tions in civilian and career astronauts
In the I4 cohort, we found no relationship between spaceflight and increased CHIP-related genetic abnormalities in the scale of three months after spaceflight.With our deep sequencing targeted probe approach ( Mean Depth ∼15,000x ) , we found putativ e m utations pr esent in subject C004's and C003 targeted samples but not C001 and C002.DNMT3A gene mutations chr2: g.25235778C > G, chr2: g.25234347G > C, chr2: g.25241591C > A are all missense variants with a likely low/moderate impact on function.A frameshift mutation, chr2: g.25246732GTCGTGGC AC ACCGGGAAC AGCTTCCCCGC > G, was detected in astronaut 004, as well as a stop-gain function mutation chr2: g.25247647G > A. Although DNMT3A presented the higher number of mutations, CHEK2 presented a missense mutation chr22: g.28734438C > T in C004, and ASXL1 presented a fr ameshift m utation c hr20: g.32434638AG > A in both C003 and C004.Ho w e v er, with a VAF of < 0.02 for all timepoints, although detectable, there is no evidence to suggest a clinical effect on the subjects , that is , spaceflight ha ving no discernible effects up to 3 months post-flight.TET2 exhibited a missense mutation chr4: g.105275662G > T at a VAF ∼ 0.3 in C003.Ho w e v er, putativ e m utations in CH-linked genes were not found at disproportionate allele frequencies post-flight, nor were de novo mutations seen in the post-flight samples.
Although our findings suggest that no pathological genomic alter ations occurr ed as a function of spaceflight in the I4 cohort, further longitudinal tr ac king of the astronauts will help us discern the long-term effects, if any, of short dur ation missions.Pr e vious studies of the relationship between clonal hematopoiesis and astronauts has elucidated that spaceflight incr eases pr oportion of CH abnormalities in subjects when compared to healthy controls [ 3 , 4 ].It is not clear why that is the case, or when in an astronaut's career these abnormalities begin to appear, a k e y limitation of studying career astronauts.Previous studies have also shown the influence of acute e v ents on the de v elopment of CH abnormalities, such as first responders from the world trade center tragedy [ 17 ].Ther efor e, the I4 mission is uniquely positioned to characterize the clonal dynamics associated with a short-duration flight, and will allow future comparisons between career astronaut CH abnormalities and civilian space tr av elers.Ther e is an opportunity to study the compound effects of spaceflight between missions on CH abnormalities if one or more of the I4 astr onauts r eturns to space .T his can be le v er a ged for a better understanding of longterm risk and clonal evolution.
The space environment is known to be chronically inflammatory as astronauts in previous studies have reported increased inflammation markers with the flight duration [18][19][20].In the NASA Twins Study, both career astronauts had increased mutational burden in epigenetic regulators such as DNMT3A and TET2 compared to prostate cancer patient controls.Although only one of the twins was in orbital spaceflight for a year, both twins displayed comparable CHIP-related mutations shortly after flight and for se v er al years of follow-up [ 6 ].We continue to monitor the CHIP-related mutational burden of both astronauts years post-flight.The VAF of mutations found in both subjects has, as expected, incr eased ov er time, and a new missense variant has been identified in subject HR in the last 3-year followup.This is to be expected as CHIP-mutational burden is known to increase as a function of aging.Both astronauts, ho w ever, presented clonal burdens that preceded their age-matched control for over two decades at the time of the original study.The fact that TW ( space subject ) did not present a greater mutational burden post-flight than HR ( ground control ) , despite the length of his last mission, might indicate that the ov er all number of flights, rather than their duration, is a greater extrinsic factor for clonal positive selection.The I4 cohort provides a unique opportunity to assess the short and long term physiological effects of a singular spaceflight mission.As civilians, the I4 cohort's physiological findings are likely a better representation of what first-time fliers may expect as we gear up toward an increased human presence in space.
Mor eov er, the longitudinal comparison of point mutations and indels from whole-genome sequencing data supports the results of the targeted deep-sequencing CHIP panel, wher ein ther e is not a significant increase in the genome-wide mutational burden for first-flight civilian astronauts .T hese results favorably indicate that short-time spaceflight does not contribute to ov er all increased genomic instability, at least on a timescale of se v er al months post-flight.Extending the analysis to future missions and continuous longitudinal tr ac king for years post-flight could shed more light on the long-term consequences of spaceflight in relation to its altitude and duration, or for low VAF alleles, as well as r epeated exposur e of astr onauts to space r adiation.
Finally, we note that transient changes in mitochondrial gene expr ession wer e seen as a r esponse to spaceflight acr oss a br oad range of immune cells within the I4 cohort.Longitudinal gene expr ession anal yses r e v ealed a conserv ed tr end acr oss both ada ptive and innate immune cells that had not been described before in the spaceflight liter atur e.In comparisons between L-92 vs R + 1 and R + 1 vs R + 82 all immune cell types demonstrated incr eased expr ession of MTRNRL12, MTRNR2L8, and MT-ND4L in R + 1. MTRNRL12 and MTRNR2L8 are pseudogenes thought to be involved in negative regulation of the execution phase of apoptosis [ 21 ].As suc h, incr eased MTRNRL12 & MTRNR2L8 expression could indicate a physiological attempt at r egulating incr eased imm une cell death.Incr eased imm une cell apoptosis was seen in our cfDNA analysis, at all post-flight timepoints, providing support to the gene expression alterations seen.
Of note, MT-CO2 and MT-CO3 gene expr ession wer e downr egulated post-flight ( R + 1 ) compared to pre-flight ( L-92 ) within CD86 + DCs , NC AM + NK cells , CD14 + monocytes , and CD3 delta + CD8 + & CD4 + T cells ( Fig. 4 B, D, F & s upplememtary Fig. S4 B, D, F, H ) .MT-CO2 and MT-CO3 encode subunits of cytoc hr ome c oxidase, which is involved in the reduction of oxygen to water.Decreases in MT-CO2 and MT-CO3 expression cause significant cytoc hr ome c oxidase and mitochondrial complex IV deficiencies, which may lead to tissue maladaptations [ 22 , 23 ].Most mitochondrial genes return to basal levels of expression in our L-92 vs R + 82 comparison, suggesting relief from spaceflight-induced mitochondrial o xidati ve stress .DCs , monocytes , and T cells showed incr eased MT-ND4L expr ession ov er two months after r eturn fr om flight.The MT-ND4L gene encodes for NADH-ubiquinone oxidoreductase chain 4L, a component of the respiratory chain Complex I, a pr otein r equir ed for electr on tr ansfer and dehydr ogenation fr om NADH to ubiquinone.Longitudinal changes in its gene expression may suggest that mitochondrial ATP production is altered long after return to earth.Ho w ever, whether the gene expression is representativ e of decr eased activity of the electr on tr ansport c hain or if it is a physiological response seeking to r estor e mitoc hondrial function due to the hypofunctional aberrant effects of o xidati ve str ess, r emains to be elucidated.
In the spaceflight liter atur e, human induced pluripotent stem cell-deri ved cardiom yocytes had increased expression of genes related to the mitochondrial electron transport chain, mitochondrial transit peptide, mitochondrial translocation, and mitoribosomes [ 24 ].The NASA Twins Study also r e v ealed tr ansient enrichment of mitochondrial signaling that returned to baseline in the r ecov ery period [ 25 ].Our findings, in contr ast, r e v eal a mor e complex picture of mitochondrial gene expression changes where some ar e consistentl y upr egulated acr oss cell types and others ar e tr ansientl y downr egulated in R + 1 but return to basal le v els in R + 82.

Immune adaptations to short duration spaceflight
Immune system dysregulation in crewmembers following spaceflight has been consistentl y r eported, r e v ealing ada ptiv e imm une system changes and general shifts to w ar ds a Th2 T cell phenotype [ 26 , 27 ].Space environment exposure leads to suppressed activ ation and r eduction in T cells, along with fluctuations in cytokine gene expr ession, suc h as increased TNF-a, interleukin-8, interleukin-1r a, thr ombopoietin, VEGF, and v arious c hemokines [ 2 , 27 , 28 ].Our PBMC phenotype analysis offers insight into alterations within immune cell subpopulations, cell state, and genotype associated with short-duration orbital flight in an all-civilian cr e w.PLCG2 is known for its role in immune cell signaling by cleaving PIP2 into IP3 and diacylgl ycer ol ( DAG ) , k e y second messenger molecules that are involved in immune and growth factor receptors [ 29 ].Additionally, although the exact physiological pathway is not known, PLCG2 has been found to positiv el y affect mitoc hondrial r espir ation [ 30 ]. Dendritic cells exhibit incr eased AHR expression in R + 1 vs L-92 ( Fig. 4 , Supplememtary Fig. S4 C ) , a transcription factor known to promote the de v elopment of pr oinflammatory Th17 and Th22 cells [ 31 , 32 ].T-helper ( CD4 + ) populations pr esent decr eased IL-32 expr ession ( Supplememtary Fig. S4 D ) in R + 82 time points r elativ e to pr e-flight L-92, suggesting the tr ansient c hanges to ada ptiv e imm une function can last up to two months, a finding concurrent with the body of literature of spaceflight r esearc h [ 33 ].
Immune dysfunction involving cytotoxic T cells is a proposed side effect of space tr av el [ 34 ], associated with an elevated risk of opportunistic infections during spaceflight, such as latent herpes virus reactivation [ 35 ].Short-term exposure to spaceflight conditions may affect innate immune system components, including functional alterations of k e y immune cell populations like CD14 + monocytes and dendritic cells.Micr ogr avity-induced c hanges can r esult in decr eased neutr ophil and macr opha ge pha gocytic activity, thus affecting the initial response to pathogens .Moreo ver, microgra vity ma y alter monocyte maturation and function, which play a critical role in antigen presentation and T cell activation.Kaur et al. ( 2005 ) describe that, although the percentage of CD14 + monocytes was not significantly reduced in astronauts following spaceflight, days after landing, the ability of these cells to engulf pathogens , degranulate , and mount an imm une r esponse was reduced compared to that of gr ound contr ols [ 36 ].Though little is known about the influence of spaceflight on dendritic cell function, their role in inflammation is well c har acterized as the y play k e y roles in modulating cytokine pr oduction, mediating migr ation, and enhancing antigen ca ptur e and processing [ 37 ].
Finally, while the long-term effects of short-duration spaceflight on the immune system and the genome ar e minimall y understood, particularly in civilians and first-fliers, these data are critical as we expand human presence in orbit and deep space.Additional studies are needed across varying cell types, mission dur ation, and cr e w bac kgr ound, whic h can also help guide countermeasures.
Of note, the short-duration space shuttle studies revealed decr eased l ymphocyte r esponse, post-flight neutr ophil incr ease, and eosinophil decr ease, corr elating with in-flight str ess r ather than micr ogr avity [ 28 , 38 , 39 ].Latent her pesvirus r eactiv ation and cytomegalovirus shedding were also observed during space shuttle flights [ 18 ].Similar adaptations occurred during ISS missions, as cr e w members experienced mild infectious diseases, atypical allergies, or dermatitis without significant operational impact [ 18 , 32 ], which seemed to be a function of mission duration.A 12-year ISS immune data comparison indicated improvements in immunity, stress, and viral reactivation due to operational and biomedical countermeasures onboard ISS, such as resupply frequency, impr ov ements in personal communication, exercise equipment, and food quality and variety, suggesting the potential for improved quality of life on the health of astr onauts [ 40 ].Suc h oper ational and biomedical considerations from earlier missions, as well as the cellular and molecular data shown here, can help future crews and missions maintain cr e w health and safety, as well as guide efforts for lifetime health studies of astronauts.[41][42][43][44][45][46][47][48][49][50][51].Select data can be visualized online through the SOMA Data Explorer: https://soma.weill.cornell.edu .The GenBank viral database used was the most recent as of 2022-07-26.The GTDB database used was the 202 release .T he MetaPhlan4 database w as mpa_vJ an21_CHOCOPhlAnSGB_202103. The Kraken2 database contained all NCBI listed taxa ( bacteria, fungal, and viral genomes ) in RefSeq, as of 2022-09-01.The Phanta database was the most recent as of 2022-08-01.The Bakta databases were the most recent as of 2022-08-18.Code used to generate Figures and analyses from this project is available at the GitHub repository for code sharing and annotation: https://github.com/eliah-o/inspiration4-omics .

Figure 1 .
Figure 1.Telomere length dynamics, mitochondrial, and cell-free DNA cell lysis Identity as a function of spaceflight.( A ) Telomere length dynamics assessed by MMqPCR in I4 cr e w members.Normalized analysis shows increased average telomere length during orbital flight compared to pre-flight baseline and in R + 1 compared to pre-flight through Tuk e y-corrected re peated measures ANOVA ( P < 0.001 ) and ( P < 0.02 ) , r espectiv el y.A r a pid decr ease in telomer e length was also seen after landing during post-flight r ecov ery months.( B ) Autosomes do not show an y spaceflight-r elated c hange in RPKM ( reads per kilobase per million reads ) , as exemplified using chr21.The cf-mtDNA is significantly more enriched in all samples-significance le v els ( P -v alue ) of the Wilcoxon test are as follows: Pre-Flight < 0.001, Post-Flight = 0.029, Recovery < 0.001.( C ) Observed cf-mtDNA levels show high inter-sample heterogeneity.( D ) Tissue of origin deconvolution for circulating cfDNA fr a gments r e v eals an incr eased cfDNA signatur e of both ada ptiv e and innate immune cells post-landing and during reco very.T he enrichment of tissue signatures from the HPM [ 13 ] was calculated based on inferred gene expression and nucleosomal footprinting of the cfDNA fragments.Average correlation coefficients ( multiplied by -1 ) over technical replicates are depicted for each sample and time point.The heatmap was subsetted to relevant tissue signatures; extended data is available in S upplememtary Fig. S1 and TableS1, see online supplementary material.( E ) Cross-examination of cfDNA origin deconvolution using each of I4 astronaut cell-subpopulation-specific expression mark ers, deri ved from peripheral blood ( PBMC ) dataset, supports the results in D. We note an incr eased pr esence of innate and ada ptiv e imm une cfDNA in R + 1 and R + 82 when compar ed to all other timepoints .T he extended data is a vailable in S upplememtary Tables S2 , see online supplementary material.

Figure 2 .
Figure 2. Targeted deep sequencing of CHIP-related genes exhibit genomic stability and comparable mutational burden as a function of spaceflight.( A ) Putati ve dri ver mutational burden of CHIP-related genes were found in two out of the four astronauts, longitudinal follow-ups demonstrate stability during the months leading and following spaceflight.( B&C ) Variant allele frequency of putative driver mutations ( chr2: g.25235778C > G, chr2: g.25234347G > C, chr2: g.25241591C > A, chr2: g.25246732GTCGTGGC AC ACCGGGAAC AGCTTCCCCGC > G, chr2: g.25247647G > A, chr22: g.28734438C > T, chr20: g.32434638AG > A, chr4: g.105275662G > T ) in subjects 003 and 004 remained comparable both during pre-flight timepoints and post-flight r ecov ery.( D&E ) Comparison of variant allele frequency of TET2.p.Cys1273Tyr mutation in TW ( spaceflight ) subject of the NASA Twins Study and variant allele frequency of DNMT3A.p.Trp698Ter and de-novo mutation DNMT3A.p.Asp856Gly in HR ( ground subject ) sho w ed a relative stable VAF during the initial stages of the study but an increased variant allele frequency through a 6 year follow-up period.

Figure 3 .
Figure 3. Whole genome sequencing and variant calling reveals genomic stability post-flight.( A ) Comparison of variant consequences before and after spaceflight of genes with the greatest mutational burden demonstrates no gene is disproportionately mutated post-flight.We report unique counts of ( variant, effect annotation ) pairs per gene of interest.( B ) De novo mutational comparison elucidates comparable mutational burdens at a genome-wide scale.Here we count variants as unique changes of reference to alternative alleles at a given position regardless of variant annotation.( C ) Variant effect annotation for variants called in both timepoints shows most mutations are in non-coding regions.Variant annotations were ranked according to the se v erity of the v ariant effect estimated by Ensembl [ 15 ].We report unique counts of ( variant, effect annotation ) pairs across the whole genome.( D ) Timepoint comparison of variant effect annotations across the whole genome shows no disproportional mutational load for variants called uniquely post-flight.Here we show the distribution for moderate to high-severity variants.
• C for 3 min; 94 • C for 15 s, 49 • C for 15 s, for 2 cycles; 94 • C for 15 s, 62 • C for 10 s, 74 • C for 15 s, 84 • C for 10 s, and 88 • C for 15 s, for 32 cycles.The melting curve was established by a 72 • C to 95• C ramp at 0.5 • C/second increase with a 30 second hold.Standard curves were prepared using human genomic DNA ( Promega Cat # G3041 ) with 3-fold dilutions ranging from 50 ng to 0.617 ng in 3 μL per dilution.Negativ e contr ols included a no-template TelG/C only and AlbU/D only, and a combined TelG/C and AlbU/D contr ol.Samples wer e normalized across plates using a human genomic DN A standar d.Each sample was run in triplicate on a 96-well plate format and r elativ e telomer e length was established using a telomere ( T ) to albumin ( A ) ratio.