Transgenic Rescue of Spermatogenesis in Males With Mgat1 Deleted in Germ Cells

MGAT1 and complex N-glycans are required for spermatogenesis and fertility. Conditional deletion of Mgat1 in spermatogonia (Mgat1 cKO) causes reduced ERK1/2 signaling and the formation of multinucleated germ cells (MNC). Here we show that glycomics analysis of N-glycans released from fixed testis sections and analyzed by MALDI imaging mass spectrometry (MALDI-IMS) revealed a loss of MGAT1 activity in all germ cells based on the accumulation of the oligomannosyl substrate of MGAT1. To determine in which type of germ cell MGAT1 is essential for spermatogenesis, we generated Mgat1 cKO males that also expressed a Mgat1-HA transgene under the control of a germ cell-specific promoter – Stra8 for spermatogonia, Ldhc for spermatocytes and Prm1 for spermatids. Males expressing each Mgat1-HA transgene were fertile, and both males and females transmitted each transgene. When Stra8-Mgat1-HA was expressed in Mgat1 cKO males, spermatogenesis was rescued based on the morphology of testis sections, the complement of N-glycans on basigin, lectin histochemistry, MALDI-IMS, and fertility. By contrast, neither Ldhc-Mgat1-HA expressed in spermatocytes, nor the Prm1-Mgat1-HA transgene expressed in spermatids rescued spermatogenesis or fertility in Mgat1 cKO males. Therefore, MGAT1 must be expressed in spermatogonia for spermatogenesis to proceed normally.


INTRODUCTION
MGAT1 is the glycosyltransferase that transfers GlcNAc to Man 5 GlcNAc 2 Asn to initiate complex N-glycan synthesis (reviewed in Stanley et al., 2015Stanley et al., -2017. We previously showed that deletion of Mgat1 in spermatogonia using Stra8-iCre prevents the formation of sperm, and thus causes infertility in male mice (Batista et al., 2012). The block in spermatogenesis is at the spermatid stage of germ cell differentiation. Spermatids lacking Mgat1 form multinuclear cells (MNC) or syncytia by ∼26 days post-partum (dpp), but the number and morphology of spermatogonia, spermatocytes and Sertoli cells appear unperturbed (Biswas et al., 2018). Here we address two questions -(1) does mis-expression of Mgat1 in spermatogonia, spermatocytes or spermatids disrupt spermatogenesis? and (2) if not, can the block in spermatogenesis in Mgat1 cKO males be rescued by expressing a Mgat1 transgene in spermatogonia, spermatocytes, or spermatids? The germ cell specific promoters that we used to express Mgat1 in each germ cell type have been characterized previously -the Stra8 promoter was used for expression in spermatogonia (Sadate-Ngatchou et al., 2008), the Ldhc promoter was used for expression in spermatocytes (Li et al., 1998), and the Prm1 promoter was used for expression in spermatids (O'Gorman et al., 1997). In this paper, we show that males expressing each Mgat1 transgene were viable and fertile. Males in which Mgat1 had been deleted in spermatogonia (Mgat1 cKO) and were also expressing a Mgat1-HA transgene in spermatocytes or spermatids exhibited the spermatogenic defects typical of Mgat1 cKO males, and were therefore not rescued by the respective Mgat1-HA transgene. However, Mgat1 cKO males expressing the Stra8-Mgat1-HA transgene were fertile, and thus rescued by the transgene.

Mice
The Transgenic Mouse Facility of the Albert Einstein College of Medicine generated the transgenic mice used in this study on a FVB/NJ background. FVB/NJ mice from Jackson Laboratories (Portland, ME, United States) were used for breeding. All mice carrying a transgene were maintained as heterozygotes. The Albert Einstein Animal Institute Committee (IACUC), using the guidelines of the NIH Office of Animal Laboratory Welfare and AAALAC, International, reviewed and approved the experiments reported here under protocol numbers 20080813, 20110803, 20140803, and 20170709 to PS. Experiments were performed in compliance with these approved protocols. Mice were sacrificed by carbon dioxide asphyxiation and cervical dislocation. Testes were dissected free of surrounding tissue and weighed. Genomic DNA was prepared from toes at 7-8 days after birth (dpp), or from tail at 10 dpp for genotyping, or extracted from liver for Southern blot analysis or PCR using a Qiagen DNeasy kit (Qiagen, Hilden, Germany). All methods were performed in accordance with the relevant guidelines and regulations approved by the Einstein IACUC. Genotyping was performed by PCR of genomic DNA using the primers shown in Supplementary  Table S1, recombinant Taq polymerase, and dNTPs from New England Biolabs Inc., Ipswich, MA, United States. Quantitative PCR (qRT-PCR) to determine transgene copy number was performed as previously described (Varshney et al., 2019) using qRT-PCR primers presented in Supplementary Table S1.

MALDI-IMS
Testes of ∼90-day mice were dissected free of surrounding tissue and fixed in 10% buffered formalin. After 48 h at room temperature (RT), formalin was removed, the testis was cut vertically, each half was put into a cassette, submerged in 70% ethanol at RT and given for paraffin embedding and sectioning to the Histotechnology and Comparative Pathology Core Facility at Albert Einstein. Sections (6 µm) of formalin-fixed, paraffinembedded (FFPE) testis were mounted on Indium-Tin-Oxide (ITO)-coated glass slides. Sections were deparaffinized in xylene three times for 5 min each, rehydrated in an ethanol/water series of 100, 95, 70% ethanol:water, and subjected to antigen retrieval in citraconic anhydride buffer at pH 3 using a steamer for 30 min (Powers et al., 2014). PNGase F Prime TM (N-Zyme Scientifics LLC, Doylestown, PA, United States; 0.1 µg/µl in HPLC-grade water) was coated on each slide using an ImagePrep Station (Bruker Daltonics Inc., Bullerica, MA, United States). Slides were incubated at 37 • C for 2 h in a humidified chamber and then dried in a vacuum desiccator for 20 min. A slide image was scanned using a flatbed scanner at 1,600 dpi resolution for use in selecting fiducial marks to ensure that the laser was in position. Slides were then coated with 2,5,dihydroxybenzoic acid (DHB) matrix (30 mg/ml in acetonitrile (AcN), 0.1% trifluoroacetic (TFA) acid and 0.1 mM NaCl) using the Imageprep station. N-glycan imaging was performed by scanning and acquiring spectra from the entire section (m/z 900-4,500) on an Ultraflextreme MALDI-TOF/TOF mass spectrometer equipped with a SmartBeam II laser at 1 kHz and 100 µm raster width. MALDI-TOF data were processed by FlexImaging software (v 4.0) to generate N-glycan ion maps. For MS/MS analysis, the N-glycans on the tissue surface were extracted with 25% AcN, 0.1%TFA, mixed with DHB matrix (12 mg/ml in 50% AcN, 5 mM NaOH), spotted on a MALDI target and analyzed using the Ultraflextreme MALDI-TOF/TOF mass spectrometer.

Western and Lectin Blot Analysis
Lysates were made from decapsulated testes of adult mice as previously described (Biswas et al., 2018). Protein was extracted in homogenization buffer (1.5% Triton X-100 in distilled water containing protease inhibitors (#05892791001, Roche Diagnostics GmbH, Mannheim, Germany) by pellet pestle (#749540000, Kontes Glass Co., Vineland, NJ, United States), centrifuged at 3,000 rpm for 10 min at 4 • C, and supernatant protein was determined by Bradford assay (Bio-Rad, Hercules, CA, United States). For detection of HA and ACTB, 40-60 µg lysate protein was separated on a 10% SDS polyacrylamide gel, transferred to nitrocellulose membrane (#162-0095, Bio-Rad) in 10% methanol, blocked in 5% non-fat milk for 1 h at RT, and incubated overnight at 4 • C with primary Ab. Membranes washed 3X in TBST and 1X in TBS, were then incubated with HRPconjugated secondary antibodies in blocking solution for 1 h at RT in the dark. For detecting BSG, or GAPDH, 40-60 µg lysate protein was separated by 10% SDS-PAGE and transferred to PVDF membrane (10% methanol, 60 min), blocked in 3% fish gelatin (#G7041, Sigma-Aldrich) in TBST, and incubated in primary Abs. After washing, membranes were probed with an appropriate HRP-conjugated secondary antibody for 1 h at RT, treated with Super Signal West Pico chemiluminescent substrate (#34080, Thermo Fisher) and exposed to film. Lectin blot analysis is described in Supplementary Material.

Statistical Analysis
All results are presented as mean ± SEM for ≥ 2 independent experiments, including replicates as noted. A one-way ANOVA or two-tailed unpaired Student's t-test with Welch's correction was used to determine p-values using Graph Pad Prism 7.0a (Graph Pad Software Inc., La Jolla, CA, United States) as noted.

RESULTS
We have previously shown that conditional deletion of Mgat1 in spermatogonia using Stra8-iCre (Mgat1 cKO) leads to a block in spermatogenesis and infertility (Batista et al., 2012;Biswas et al., 2018). The loss of MGAT1 occurred in all testicular germ cells based on imaging of testis sections using MALDI-Imaging mass spectrometry (MALDI-IMS) ( Figure 1A). The accumulation of the substrate of MGAT1 (Man 5 GlcNAc 2 Asn) and the marked reduction of complex N-glycans in the mutant is apparent in the density maps for individual N-glycans ( Figure 1A) and in the MS profiles (Figures 1B,C). Consistent with this is the fact that the three complex N-glycans of basigin (BSG), a known target of MGAT1 in Mgat1 cKO testis, were removed by endoglycosidase H (Endo H), which cleaves oligomannosyl but not complex N-glycans (Figure 1D), as shown previously (Batista et al., 2012).
To determine the germ cell type that requires complex N-glycans for survival or differentiation, we used a transgenic rescue strategy. A Mgat1-HA cDNA was cloned under the control of the Stra8 promoter from Stra8-iCre (Sadate-Ngatchou et al., 2008), or the Ldhc promoter from Ldhc-LacZ (Li et al., 1998;Goldberg et al., 2010), or the Prm1 promoter from Prm1-Cre (O'Gorman et al., 1997; Figure 2A). Several transgenic founders were obtained following injection of each cDNA into fertilized eggs (5 for Stra8-Mgat1-HA, 4 for Ldhc-Mgat1-HA, and 16 for Prm1-Mgat1-HA). However, each transgene was transmitted through the germline by only one founder, respectively. Transgene expression was characterized by histology, lectin histochemistry and western blot analysis for HA and MGAT1. Testis sections from each transgenic strain stained with H&E exhibited no differences in spermatogenesis from non-transgenic control sections ( Figure 2B). Lectin blot analysis revealed increased signal for the plant lectins phytohemagglutinin-L (L-PHA) and Griffonia simplicifolia II (GSA II) in sections from mice expressing each transgene compared to non-transgenic controls ( Figure 2B). Both lectins bind to complex N-glycans and binding was increased as expected due to expression of the Mgat1-HA transgene. Western blot analysis for expression of HA or MGAT1 in testis lysates showed that the signal for each was markedly increased in Stra8-Mgat1-HA and Ldhc-Mgat1-HA transgenic males (Figures 2C,E,F) compared to non-transgenic littermates or wild type males. The HA signal of the Prm1-Mgat1-HA transgene was also increased in both testis and germ cell lysates ( Figure 2D). The Prm1-Mgat1-HA strain was investigated in additional detail. Southern analysis showed that the transgene was inserted as a linear concatemer (Supplementary Figure S1). Quantitative PCR of genomic DNA from liver revealed that the Prm1 promoter and the Mgat1 gene were present in ∼8-11 copies in the three mice analyzed, compared to two copies of each in non-transgenic liver DNA (Supplementary Material). Lectin blot analysis showed a large increase in the binding of L-PHA to glycoproteins in Prm1-Mgat1-HA testes, consistent with increased expression of complex N-glycans due to the transgene (Supplementary Figure S2). By contrast, binding of Galanthus nivalis lectin (GNA) which recognizes oligomannosyl N-glycans was very low in proteins from transgenic testes, again consistent with high conversion of oligomannosyl to complex N-glycans by MGAT1. Proteins from Mgat1 cKO testes or Lec1 CHO cells that lack complex N-glycans also bound GNA well, but L-PHA poorly (Chen and Stanley, 2003;Huang and Stanley, 2010). Nontransgenic testis extract and wild type CHO cells bound L-PHA well, and GNA poorly.
The fertility of each of the transgenic strains was similar to non-transgenic littermates and each transgene was transmitted ∼50% of the time. Thus, there were no apparent deleterious consequences of mis-expressing or over-expressing To investigate whether the respective transgenes could rescue spermatogenesis in a Mgat1 cKO male, each transgene was crossed onto Mgat1[F/F] males, which were then crossed with Mgat1[F/F]:Stra8-iCre females, to generate Mgat1 cKO-Stra8-Mgat1-HA, Mgat1 cKO-Ldhc-Mgat1-HA or Mgat1 cKO-Prm1-Mgat1-HA "rescue" males, respectively. Several "rescue" males for each transgene were mated to wild type FVB females and the progeny were genotyped to determine whether Mgat1 had been deleted by Stra8-iCre, and to show that the respective Mgat1-HA transgene was present. Six Mgat1 cKO males expressing the Stra8-Mgat1-HA transgene were fertile and gave an average of 7 pups per litter from 10 litters. Histological analysis revealed ∼13% testis tubules contained MNCs and ∼3% epididymi contained cellular debris amongst their sperm ( Figure 3A). However, the majority of testis tubules were rescued. Thus, lectin histochemistry showed that complex N-glycans bound by the FIGURE 2 | Generation and characterization of Mgat1 transgenic mice. (A) Schematic of cDNA constructs used to generate transgenic mice. Expression of Mgat1, with a C-terminal HA tag (Mgat1-HA) was driven by germ cell specific promoters Stra8, Ldhc and Prm1, respectively. The primers for genotyping spanned the promoter and cDNA sequences so that a transgene-specific region was amplified (Supplementary Table S1). (B) Top panels. Representative images of H&E staining of adult non-transgenic (Control) and transgenic testis sections at 28 dpp. Scale bar 50 µm. No morphological changes were observed for Mgat1 mis-expressed in spermatogonia, spermatocytes, or spermatids at 28 dpp (8-12 animals per transgene were analyzed; 100 tubules counted per section, two sections per mouse). Middle panels. Binding of L-PHA to testis sections from 28 dpp control males (n = 3, two sections per mouse), and transgenic males (n = 4-6, two sections per mouse). Scale bar 100 µm. Bottom panels. Binding of GSA II to testis sections from 28 dpp control (n = 3) and transgenic (n = 4-6) males. Scale bar 100 µm. (C) Western blot analysis of HA expression in testis extracts from control and transgenic mice with expression in spermatogonia or spermatocytes. (D) Western blot analysis for HA in left and right testis extracts from two Prm1-Mgat1-HA males and a non-transgenic male. Extracts of germ cells purified from a transgenic and a non-transgenic male are also shown. (E,F) Western blot analyses of Stra8-Mgat1-HA and Ldhc-Mgat1-HA transgenic testis lysates for expression of MGAT1 compared to FVB controls. The ACTB control for (C) is the same for (E). *Non-specific loading control. lectins L-PHA and GSA II were present in Mgat1 cKO-Stra8-iCre-Mgat1-HA transgenic mice, but absent from Mgat1 cKO control ( Figure 3B). Western analysis showed the expression of the transgene was robust ( Figure 3C) and that BSG N-glycans were ∼70% resistant to Endo H digestion, consistent with the presence of mainly complex N-glycans ( Figure 3D). MALDI-IMS of testis sections showed a marked increase in biantennary complex N-glycans in Mgat1 cKO-Stra8-Mgat1-HA males compared to Mgat1 cKO males (Figures 3E,F). The latter exhibited the characteristic accumulation of MGAT1 substrate. Thus, the Stra8-driven transgene substantially rescued the deletion of Mgat1 in the majority of testis tubules leading to apparently normal fertility.
By contrast, 10 Mgat1 cKO-Ldhc-Mgat1-HA transgenic males and 4 of 6 Mgat1 cKO-Prm1-Mgat1-HA males gave no progeny when housed with wild type females, several of which were known to be fertile. This was despite robust expression of the Mgat1-HA transgene in each strain ( Figures 4A,B). Males that gave no progeny exhibited the typical disruption of spermatogenesis of Mgat1 cKO males (Batista et al., 2012;Biswas et al., 2018) with many MNCs per tubule, no sperm and debris in the epididymis (Figure 4A and Supplementary Figure S3). The N-glycans of BSG were fully sensitive to removal by Endo H in these males (Figures 4C,D), showing there were no complex N-glycans on BSG. The two Prm1-Mgat1-HA males that did not exhibit 100% deletion of the Mgat1 gene, had slightly rescued spermatogenesis and were fertile (Supplementary Figure S3), apparently due to MGAT1 activity derived from the Mgat1[F] alleles observed in progeny, and not to rescue by the Prm1-Mgat1-HA transgene. In the final analysis, only the Stra8-Mgat1-HA transgene rescued the fertility of Mgat1 cKO males.

DISCUSSION
Spermatogenesis requires the synthesis of several glycan species, including complex N-glycans (Batista et al., 2012;Biswas et al., 2018;Akintayo and Stanley, 2019). However, determining mechanistic bases for these requirements is a challenge since glycans are expressed on many different glycoproteins and glycolipids. We have taken a combined genetic/molecular/biochemical approach to identify features of the requirement for complex N-glycans during spermatogenesis. We showed that deletion of Mgat1 at ∼3 dpp in male germ cells manifests morphologically at about 26 dpp as fusion of spermatids (Biswas et al., 2018). However, molecular changes can be observed earlier. Microarray analyses of transcripts from male germ cells of 22 or 23 dpp mice revealed gene expression differences consistent with the premature activation of spermatogenesis and spermiogenesis genes in Mgat1 cKO germ cells, and implicated reduced EGF/PDGF receptor signaling. Indeed, there was a marked reduction in pERK1/2 in germ cells of Mgat1 cKO males compared to controls. In this paper, we continue the investigation into roles for complex N-glycans in spermatogenesis by asking in which germ cells MGAT1 is required for spermatogenesis to proceed normally.
To address this question, we engineered transgenic mice expressing a Mgat1-HA cDNA in spermatogonia, or spermatocytes, or spermatids with the aim of determining which, if any, could rescue the defects in Mgat1 cKO spermatogenesis.
First, we investigated whether mis-expression of Mgat1-HA in specific germ cells would disrupt their function. We found that spermatogenesis and fertility were similar to non-transgenic controls in each transgenic strain. That is, we demonstrated that mis-or over-expression of Mgat1 in spermatogonia or spermatocytes or spermatids did not apparently affect spermatogenesis, and did not significantly alter fertility. Thus, we could investigate whether one or more of the transgenes would rescue the spermatogenic defects in Mgat1 cKO mice. We report here that the only transgene to rescue Mgat1 cKO males was the Mgat1-HA expressed in spermatogonia under the Stra8 promoter. The degree of rescue was high but not complete, as shown by the presence of ∼13% testicular tubules that contained MNCs, epidydimi with pyknotic cells in the lumen, the presence of an increased amount of Man 5 GlcNAc 2 Asn in germ cells by MALDI-IMS, and the fact that the N-glycans of BSG had an increased proportion of oligomannosyl N-glycans sensitive to Endo H. Nevertheless, Mgat1 cKO-Stra8-Mgat1-HA males were as fertile as controls. By contrast, expression of Mgat1-HA in spermatocytes under the Ldhc promoter, or spermatids under the Prm1 promoter, did not rescue the Mgat1 cKO phenotype. These findings are striking and point to spermatogonia as being the key germ cell in which complex N-glycans must function for spermatogenesis to proceed. It is interesting that the MNCs observed in the absence of MGAT1 are formed by round and elongated spermatids (Batista et al., 2012;Biswas et al., 2018). Spermatogonia and spermatocytes appear normal, and are present in their usual numbers, when Mgat1 is deleted in spermatogonia (Biswas et al., 2018). However, the increased level of apoptosis present in Mgat1 cKO germ cells (Batista et al., 2012) probably affects primarily spermatocytes. In addition, the molecular changes observed at 22 or 23 days after birth (Biswas et al., 2018), largely before spermatids are present. presumably lead to the dramatic fusion of spermatids that begins at ∼26 days after birth.
Based on the importance of MGAT1 in mouse spermatogenesis and fertility (Batista et al., 2012;Biswas et al., 2018), it is to be expected that mutations in the MGAT1 gene are a source of infertility in men. However, a recent comprehensive review of mutations associated with infertility in men did not report any with a mutation in the MGAT1 gene (Oud et al., 2019). Nevertheless, our studies suggest that MGAT1 is a strong candidate and will likely be revealed as a basis for infertility in future genomic studies. MGAT1 is also a potentially excellent target for a male contraceptive strategy. A small molecule inhibitor or a peptide (Huang and Stanley, 2010) that inhibits MGAT1 activity would ostensibly have few effects on spermatogonia or spermatocytes (Biswas et al., 2018), but would induce fusion of spermatids and prevent the formation of sperm (Batista et al., 2012), thus having contraceptive activity. Moreover, these effects could be reversible, since spermatogonia are not observably impacted by inhibition of MGAT1 activity (Biswas et al., 2018).

DATA AVAILABILITY STATEMENT
All datasets generated for this study are included in the article/Supplementary Material.

ETHICS STATEMENT
Animal experiments were reviewed and approved by the Einstein Institutional Animal Care and Use Committee under protocol numbers 20110803, 20080813, 20140803, and 20170709. AUTHOR CONTRIBUTIONS BB generated transgenic and rescue mice, characterized them by histology, lectin histochemistry, western blotting, and fertility and co-wrote the manuscript. FB generated and characterized the Prm1 transgenic and rescue lines and read the manuscript. AA bred experimental mice, characterized rescue mice, and edited the manuscript. JA performed and interpreted MALDI-IMS and edited the manuscript. PS conceived and guided the experiments, acquired funding, analyzed data, and co-wrote the manuscript.

FUNDING
This work was supported by a grant from the National Institute of General Medical Sciences RO1 GM 105399 to PS, by a Lalor Foundation Fellowship to BB, and in part by the Albert Einstein Cancer Center grant from the National Cancer Institute PO1 CA 13333 which supports the Core Facilities and PS.