Transcriptomic landscape and potential therapeutic targets for human testicular aging revealed by single-cell RNA sequencing

Background: Testicular aging is known to cause male age-related fertility decline and hypogonadism, but the underlying molecular mechanisms remain unclear. Methods: We survey the single-cell transcriptomic landscape of testes from young and old men and examine age-related changes in germline and somatic niche cells. Results: In-depth evaluation of the gene expression dynamics of germline cells reveals that disturbance of base-excision repair pathway is a major feature of aging spermatogonial stem cells (SSCs), suggesting that defective DNA repair of SSCs may serve as a potential driver for increased de novo germline mutations with age. Further analysis of aging-associated transcriptional changes shows that stress-related changes and apoptotic signaling pathway accumulate in aged somatic cells. We identify age-related impairment of redox homeostasis in aged Leydig cells and find that pharmacological treatment with antioxidants alleviate this cellular dysfunction of Leydig cells and promote testosterone production. Lastly, our results reveal that decreased pleiotrophin (PTN) signaling is a contributing factor for testicular aging. Conclusions: These findings provide a comprehensive understanding of the cell-type-specific mechanisms underlying human testicular aging at a single-cell resolution, and suggest potential therapeutic targets that may be leveraged to address age-related male fertility decline and hypogonadism. Funding: This work was supported by the National Key Research and Development Program of China (2018YFA0107200, 2018YFA0801404), the National Natural Science Foundation of China (32130046, 82171564, 82101669, 81871110, 81971759), the Key Research and Development Program of Guangdong Province (2019B020234001), the Natural Science Foundation of Guangdong Province, China (2022A1515010371), the Major Project of Medical Science and Technology Development Research Center of National Health Planning Commission, China (HDSL202001000), the Open Project of NHC Key Laboratory of Male Reproduction and Genetics (Family Planning Research Institute of Guangdong Province) (KF202001), the Guangdong Province Regional Joint Fund-Youth Fund Project (2021A1515110921), the China Postdoctoral Science Foundation (2021M703736).


Aging-related cellular alterations along the trajectories of spermatogenesis 202
To identify germline changes during aging, we performed a focused analysis of germ cell 203 clusters. We first investigated the proportions of germ cells at each stage. We identified the 204 following five broad developmental stages of germ cells based on their marker genes: SSCs 205 (UTF1+, ID4+), Diff-SPGs (STRA8+, DMRT1+), SPCs (SYCP1+, SYCP3+, and DAZL+), 206 RSs (SPAG6+, CAMK4+), and ESs (PRM3+, HOOK1+, and TNP1+) (Figures 2A, 2B). We 207 further inferred the trajectories of spermatogenesis by performing an orthogonal pseudotime 208 analysis using the Monocle package (Qiu et al., 2017). This analysis revealed that complete 209 spermatogenesis was present in all samples ( Figure 2C, 2D). We then parsed out the germ cells 210 by developmental stage to examine their relative compositions at different ages. Interestingly, 211 these percentages were not significantly different from SSCs to ESs. In contrast, the proportions 212 of germ cells at RS stage and beyond tended to decline with age ( Figure 2E). 213 Immunofluorescence analysis of the germ cell marker, DEAD-box helicase 4 (DDX4), 214 supported the notion that germ cells were reduced in the testes of the old group ( Figure 2F). 215 Further analysis revealed that the numbers of UCHL1+ SSCs were equivalent between the 216 young and old groups ( Figure 2G), indicating that a change in the quantity of SSCs might not 217 be largely involved in the aging-associated deficiency of spermatogenesis. The number of 218 peanut agglutinin (PNA)+ RSs and ESs was significantly lower in testes of young versus elderly 219 individuals ( Figure 2H). This indicated that germ cell differentiation was severely impacted by 220 aging, which is consistent with our histological examinations ( Figure 1A). 221 Taken together, our transcriptomic data combined with the results of our immunofluorescent 222 studies indicate that there are cellular differences in spermatogenesis-related cells of the young 223 and old groups, and that RSs and ESs decreased considerably with age.

247
Aging-related molecular alterations along the trajectories of spermatogenesis 248 We next focused on age-related transcriptional changes in the identified germline cell types. 249 Aging has been associated with increased transcriptional noise (Nikopoulou,Parekh,& Tessarz,250 2019). Here, calculation of the age-relevant coefficient of variation (CV) revealed that SPCs, 251 Diff-SPGs, and SSCs exhibited higher transcriptional noise than later-stage germline cells 252 ( Figure 3A), indicating that aging caused higher variability in early-stage germline cells  253   15   compared to late-stage germline cells. We further identified 174, 102, 90, 644, and 214  254   upregulated genes and 112, 46, 127, 891, and 665 downregulated genes (|avg_logFC| > 0.25  255 and p-value < 0.05) in the SSC, Diff-SPG, SPCs, RS, and ES germline subtypes, respectively, 256 in the old versus young groups ( Figure 4B; Supplementary file 2). Notably, only ~22% of the 257 differentially expressed genes (DEGs) were shared by at least two cell populations; the majority 258 of DEGs were cell-type-specific, indicating that aging has stage-specific effects in this setting.  Figure 3C). Therefore, we performed gene set-267 score analysis for the BER and NER signaling pathway of SSCs in the young and old groups 268 (Supplementary file 4). We observed a prominent decrease of the gene-set score of the BER 269 pathway in aged SSCs, compared to young SSCs, whereas that of the NER pathway did not 270 significantly differ between the two groups ( Figure 3D). Moreover, we found that a number of

290
Color keys, ranging from gray to orange to red, indicate the absolute value of the normalized enrichment 291 score (NES).

292
(D) Gene set score analysis for the BER and NER signaling pathways of SSCs from the young and old groups.

293
Wilcoxon rank sum test was used; p-value is indicated.

Changes in the transcriptional profiles of somatic cells during human testicular aging 312
We next quantified the populations of the main somatic cell types, including LCs, SCs, and 313 PTMs, in young and old testes by immunofluorescence analysis (Figure S4A-C). To further 314 explore the mechanisms of testicular aging at the cellular level, we compared the gene 315 expression patterns in somatic cells between the groups. By calculating the age-relevant CV(S. 316 Wang et al., 2020), we found that the aging-accumulated transcriptional noise was higher in 317 LCs, PTMs, and SCs compared to the other somatic cell types ( Figure 4A). This suggested that 318 LCs, PTMs, and SCs may be more vulnerable to age-related stress than the other somatic cell 319 types. We next explored aging-associated DEGs in somatic cells between the groups. We To identify DEGs that constantly increased or decreased in somatic cells during aging, we 328 aligned datasets of individual samples by their chronological age and explored the expression 329 patterns of 2594 DEGs that were upregulated with age and 1953 DEGs that were downregulated 330 with age in six somatic cell types ( Figure 4C). GO enrichment analysis revealed that the genes 331 upregulated with age were mainly associated with cytokine and stress responses, unfolded 332 proteins, and apoptotic signaling. In comparison, the genes downregulated with age were 333 mainly related to peptide secretion, hormone responses, and growth ( Figure 4D; Supplementary 334 file 3). As the senescence-associated secretory phenotype (SASP) is a common feature of 335 senescent cells and usually contributes to a low-grade inflammatory state (Baker & Petersen,336 2018), we questioned whether the somatic niche in the aged testis might present an elevated 337 20 SASP environment. Indeed, our analysis revealed that somatic cells exhibited much higher 338 gene-set scores for SASP ( Figure 4E). Notably, LCs, PTMs and Ms showed elevated SASP 339 scores with age ( Figure 4F, 4G; Supplementary file 4), indicating that these three cell types had 340 strong contributions to the age-related inflammatory state in human testis. 341 Next, we performed an integrative comparative analysis of aging-associated DEGs with 342 aging/longevity-associated genes from the GenAge database (Tacutu et al., 2018) and further 343 found that many aging/longevity-associated genes were differentially expressed in one or more 344 somatic cell type ( Figure S4D        LCs produce testosterone and are thus critical for reproductive function and general 382 health (Zirkin & Papadopoulos, 2018). LC dysfunction causes testosterone deficiency, arrested 383 spermatogenesis, and infertility. Here, we observed the highest level of aging-accumulated 384 transcriptional noise in LCs ( Figure 4A). Thus, we next focused on the aging-associated 385 changes of gene expression in LCs. For functional validation, we assessed the ability of aged 386 LCs to produce testosterone. We isolated primary LCs from human testicular tissues of both 387 groups ( Figure S5A, B). Assessment of testosterone levels in the media revealed that LCs from 388 the young group produced significantly more testosterone than those from the old group (   Figures 5H, S5C). Interestingly, however, the antioxidant treatments 430 recovered testosterone production in isolated primary LCs, whereas the vehicle control did not 431 26 ( Figure 5I). The antioxidant treatments also restored testosterone production in aged human 432 testicular samples ( Figure 5J), suggesting that antioxidative strategies may rejuvenate LCs and 433 recover testosterone production in elderly humans. Collectively, these findings suggest a new 434 platform for uncovering potential intervention targets and compounds for alleviating late-onset 435 hypogonadism and testicular aging.           increasing interactions in old testis compared to young testis ( Figure 6B). Consistently, 487 investigation of the outgoing and incoming signaling patterns revealed that there were relative 488 declines of the outgoing interaction strengths for aged LCs and PTMs compared to their young 489 counterparts ( Figure 6C). These results encouraged us to further study the potential role of 490 LCs and PTMs in testicular aging. 491 When we compared the information flow for each signaling pathway sent by LCs or PTMs 492 between the young and old groups, we found that the pleiotrophin (PTN) signaling pathway 493 ranked highest in flow for young testis and declined prominently during aging ( Figure 6D). 494 Based on this finding, we investigated the PTN signaling pathway network of testicular cells. 495 In the young group, PTN signaling was mainly sent by SCs, LCs, and PTMs and received by 496 SSCs, Diff-SPGs, SPCs, and somatic cells. In the old group, in comparison, PTN signaling was 497 dramatically lower between LCs, PTMs and other cells, but relatively similar between SCs and 498 other cells ( Figure 6E). 499 In the testis, growth factors and morphogen signals play important roles in spermatogenesis. 500 We next inferred the age-related changes in the communication networks of target germ cells. 501 Similar to the above-described findings, PTN was prominent among the incoming signal 502 pathways of SSCs, Diff-SPGs, and SPCs ( Figure S6A). This supports the idea that PTN 503 31 signaling forms a potential regulatory axis between somatic cells and spermatogenesis. Our 504 analysis identified five significant ligand-receptor pairs for PTN signaling: PTN-NCL, PTN-505 SDC1, PTN-SDC2, PTN-SDC3, and PTN-SDC4. Among them, PTN-NCL contributed most 506 highly to PTN signaling in both young and old groups ( Figure S6B, S6C). The expression level 507 of PTN was downregulated in LCs and PTMs of the old group compared to the young group, 508 whereas the expression levels of NCL, SDC1, SDC2, SDC3, and SDC4 did not differ with age 509 in most of the other tested cell populations ( Figure 6F). This further supports the importance of 510

Discussion 544
In this report, we present the single-cell survey of human testicular aging, providing insights 545 into the mechanisms by which human testes age. Our analyses provide four noteworthy 546 contributions. First, we elucidated the gene expression signatures of 11 types of human 547 testicular cells (including germline and somatic cells) and identified several previously 548 unreported cell type-specific markers. Second, analysis of age-associated gene expression 549 changes revealed that DNA repair genes are downregulated in aged human SSCs, which may 550 represent a driver for the age-related increase of de novo germline mutations in males. Third, 551 human LCs exhibited aging-associated upregulation of oxidative stress genes, and antioxidant 552 treatment recovered the functions of aged human LCs. Fourth, our results suggested that 553 decreases in PTN signaling contribute to testicular aging. Together, these observations provide 554 novel insights into human testicular aging and identify new potential targets for treating human 555 disorders associated with testicular aging. Intriguingly, we also observed an elevated inflammatory response in aged LCs, which may 619 38 reflect an accumulation of ROS-induced macromolecule damage (Forman & Zhang, 2021). 620 Given that inflammation is an adaptive response to noxious stress or malfunction, the elevated 621 inflammation we observed in aged LCs implies that intracellular homeostasis becomes skewed 622 toward a chronic stress state with age. Recent evidence suggests that oxidative stress may be 623 linked to LC dysfunction and hypogonadism in rodents (Cao et al., 2004;H. Chen et al., 2005). 624 Consistently, we found that DNA oxidation markers were increased in aged LCs compared to 625 young LCs. Moreover, isolated primary human LCs from the old group generated more 626 intracellular ROS than LCs from young group, supporting the notion that oxidative stress 627 induces LC dysfunction in aging human testes. 628 Based on our present findings, we speculated that oxidative stress could be targeted as a 629 therapeutic avenue to prevent LC dysfunction in aging males. A number of reports have 630 examined the beneficial effects of antioxidants on the function of murine LCs, but mostly under 631 pathological conditions (e.g., testicular torsion or diabetes) or following exposure to various 632 toxic agents(H. Chen et al., 2005 The old donors were confirmed to have offspring, which was taken as indicating that they had 673 normal reproductive function when young. Informed consent was obtained from all of the 674 above-listed patients. 675

Tissue processing 676
After being collected from the operating room, the samples were transported to the laboratory 677 on ice in storage solution (Miltenyi Biotec, Shanghai, China) within 1 h. The tunica was 678 removed and testicular tissues were minced and washed three times with phosphate-buffered 679 saline (PBS) to eliminate the storage solution and blood. To ensure accuracy and stability, ~200 680 mg of tissue was immediately applied for scRNA-seq. Thereafter, tissue samples (~500 mg) 681 were fixed with 4% paraformaldehyde (PFA; Thermo Fisher Scientific, Wilmington, DE, USA) 682 for histochemistry or immunostaining analyses. The remaining testis tissues were 683 cryopreserved for functional assessments, as previously described (Guo et al., 2018). 684

Sample preparation for scRNA-seq 685
For single-cell sequencing, testicular samples were minced and subjected to a standard two- The RNA-seq sequencing and processed data reported in this paper have been deposited in the 938 Genome Sequence Archive (GSA for Human) with project number HRA002349. 939 The following dataset was generated: 940

Author(s) Year Dataset title Dataset URL Database and Identifier
Kai Xia 2022 Single-Cell