PM2.5 Exposure of Mice during Spermatogenesis: A Role of Inhibitor κB Kinase 2 in Pro-Opiomelanocortin Neurons.

BACKGROUND: Epidemiological studies have shown that exposure to ambient fine particulate matter with aerodynamic diameter less than or equal to 2.5 μm (PM2.5) correlates with a decrease in sperm count, but the biological mechanism remains elusive.
OBJECTIVES: This study tested whether hypothalamic inflammation, an emerging pathophysiological mediator, mediates the development of lower epididymal sperm count due to PM2.5 exposure.
METHODS: Inhibitor κB kinase 2 (IKK2) was conditionally knocked out either in all neurons or subtypes of hypothalamic neurons of mice. Effects of concentrated ambient PM2.5 (CAP) exposure on hypothalamic inflammation, the hypothalamic-pituitary-gonadal (HPG) axis, and epididymal sperm count of these mouse models were then assessed. Furthermore, to test whether hypothalamic inflammation is sufficient to decrease sperm production, we overexpressed constitutively active IKK2 (IKK2ca) either in all neurons or subtypes of hypothalamic neurons and assessed hypothalamic inflammation, the HPG axis, and sperm production of these overexpression mouse models.
RESULTS: CAP-exposed wild-type control mice vs. filtered air (FA)-exposed wild-type control mice had a higher expression of hypothalamic inflammatory markers, lower functional indexes of the HPG axis, and a lower epididymal sperm count. In contrast, all these measurements for CAP- vs. FA-exposed mice deficient of IKK2 in all neurons were comparable. We also found that overexpression of IKK2ca in either all neurons or pro-opiomelanocortin (POMC) neurons only, but not in Agouti-related protein (AgRP) neurons only, resulted in lower functional indexes of the HPG axis and a lower epididymal sperm count. Moreover, we showed that CAP- vs. FA-exposed mice deficient of IKK2 in POMC neurons had a comparable expression of hypothalamic inflammatory markers, comparable functional indexes of the HPG axis, and a comparable epididymal sperm count. DISCUSSION: This mouse model study shows a causal role of IKK2 of POMC neurons in the development of lower epididymal sperm count due to PM2.5 exposure, providing a mechanistic insight into this emerging pathogenesis. https://doi.org/10.1289/EHP8868.

Introduction

A worldwide decline in the sperm count has been noted over the past decades (Levine et al. 2017). However, the causes for this decline are not yet fully understood. The widespread and continuing pattern of this decline suggests that it is attributed primarily to adverse environmental exposures (Levine et al. 2017). It was estimated that 91% of the world population are now living in places where the fine particulate matter with aerodynamic diameter less than or equal to () level exceeds the guideline limit of the World Health Organization (Cohen et al. 2017). Mounting epidemiological studies showed that the exposure level was associated with the sperm count (Carré et al. 2017; Lafuente et al. 2016). All these findings suggest that pollution may be one of the primary culprits for the worldwide decline in sperm count and thus warrant a focused scientific effort to unravel the underlying biological mechanism.

Animal model studies are essential for delineating biological mechanisms. In line with the above-mentioned epidemiological studies (Carré et al. 2017; Lafuente et al. 2016), almost all published animal model studies demonstrated that exposure disrupted spermatogenesis and reduced epididymal sperm count (Cao et al. 2015; Qiu et al. 2018; Wei et al. 2018; Yang et al. 2019; Zhou et al. 2019). Of them (Cao et al. 2015; Qiu et al. 2018; Wei et al. 2018; Yang et al. 2019; Zhou et al. 2019), several studies showed that the disruption of spermatogenesis by exposure was accompanied by a marked testicular inflammation and/or its associated oxidative stress and impairment of the blood–testis barrier (Cao et al. 2015; Wei et al. 2018; Zhou et al. 2019), implicating a testicular inflammatory response to exposure in the disruption of spermatogenesis by exposure. However, considerable studies revealed that exposure disrupted spermatogenesis but did not elicit testicular inflammation (Qiu et al. 2018; Yang et al. 2019), suggesting that there may also be a testicular inflammation-independent mechanism for the impairment of spermatogenesis and thus sperm count by exposure.

Spermatogenesis is intricately regulated by the hypothalamic–pituitary–gonadal (HPG) axis, and suppression of the HPG axis is well known to result in diminution of spermatogenesis (McBride and Coward 2016; Nargund 2015). Notably, exposure was shown to suppress the HPG axis in mice (Qiu et al. 2018), although the causal role of this suppression in the disruption of spermatogenesis by exposure remains to be determined. Furthermore, exposure in mice was shown to elicit a marked hypothalamic inflammation (Qiu et al. 2018; Ying et al. 2014), an emerging pathophysiological mediator that is known to markedly affect the HPG axis (Lainez and Coss 2019b), and mounting studies revealed that this hypothalamic inflammation was independent of the pulmonary and subsequent systemic inflammation (Chen et al. 2018; Liu et al. 2014; Qiu et al. 2018; Sun et al. 2018; Ying et al. 2014) that is believed to mediate the induction of extra-pulmonary inflammations, including the testicular inflammation, by exposure (Rajagopalan et al. 2018). Collectively, these studies suggested that exposure might disrupt spermatogenesis and thus sperm count through induction of hypothalamic inflammation and subsequent suppression of the HPG axis, providing a potential testicular inflammation-independent mechanism for the impairment of spermatogenesis and the sperm count by exposure. Inhibitor kinase 2 (IKK2) phosphorylates inhibitor proteins and subsequently activates nuclear and thus plays a pivotal role in various inflammatory responses (Schnappauf and Aksentijevich 2020). We previously showed that neural deficiency of IKK2 blocked exposure-induced hypothalamic inflammation (Chen et al. 2018). To prove this potential mechanism, this study manipulated CAP exposure-induced or spontaneous hypothalamic inflammation via hypothalamic neuron-specific knockout (KO) or overexpression of IKK2 in mice and then assessed their effects on murine sperm production.

Materials and Methods

Animals

University of Maryland, Baltimore (UMB), is an AAALAC-accredited institution. All the animal-related procedures were approved by the institutional animal care and use committee of UMB. Nestin-Cre, POMC-Cre, and AgRP-Ires-Cre transgenic mice were obtained from Jackson Laboratories. The specificity and efficiency of POMC-Cre and AgRP-Ires-Cre were previously shown (Balthasar et al. 2004; Tong et al. 2008). mice were gifted to Z.Y. from Dr. M. Karin (generation of these mice is described in Li et al. 2003) and maintained at the University of Maryland School of Medicine. They were then back-crossed with C57Bl/6J mice for six generations. Floxed constitutively active IKK2 transgenic () mice in C57Bl/6J genetic background were generated as previously described (Otero et al. 2012). and littermate mice (8–11 wk old, ) were generated by out-crossing of and mice and consecutively exposed to filter air (FA) or concentrated ambient (CAP) for 16 wk, a duration previously shown to be sufficient to impair murine sperm production (Qiu et al. 2018). The conditional knockout of IKK2 in was previously confirmed (Chen et al. 2018). and littermate mice (8–10 wk old, ) were generated by out-crossing of and mice. The conditional KO of IKK2 in was confirmed by polymerase chain reaction (PCR) using the following primers: IKK2gs1: TAG TCC AAC TGG CAG CGA ATA C; IKK2gAs: CGC CTA GGT AAG ATG GCT GTC T; IKK2gS2: GTG GTC ATA GGT CTG GTT GTC C (Figure S1A).

(20–26 wk old, ), (18–23 wk old, ), (18–23 wk old, ), and their littermate controls were generated by out-crossing of each Cre line and mice. The overexpression of IKK2 in these transgenic mouse models is shown in Figure S1C. Briefly, (8 wk old, ), (8 wk old, ), (11–13 wk old, ), (8 wk old, ) were euthanized with 200 mg/kg pentobarbital euthanasia solution. The hypothalamus was isolated using incision sites as follows: rostral border of the optic chiasm, caudal border of the mammillary body, ventral border of the anterior commissure and lateral borders of the tuber cinereum and mammillary body complexes. The hypothalamic proteins were extracted using M-PER Mammalian Protein Extraction Reagent (Thermo Fisher Scientific) per the manufacturer’s instruction. The concentrations of proteins were determined using the BCA protein assay kit (Thermo Fisher Scientific). In addition, proteins were resolved with 10% SDS-PAGE and transferred to PVDF transfer membrane (Thermo Fisher Scientific). Immunostaining was performed using standard techniques with 1:500 anti-HPRT antibody (ab10479; Abcam), 1:500 anti-IKK2 (No. 2684; Cell Signaling Technology), and 1:1,000 HRP-linked anti-rabbit IgG antibody (No. 7074, Cell Signaling Technology). Signals were detected using supersignal chemiluminescence (Thermo Fisher Scientific) and Amersham Imager 600 (GE Health Care). We previously showed that a group size of seven was sufficient to observe a significant effect of CAP exposure on mouse sperm production (Qiu et al. 2018). To avoid cage effects, the KO or transgenic mice and their littermate controls were housed in the same cage. All the animals were euthanized using intraperitoneal injection of 200 mg/kg sodium pentobarbital euthanasia solution. The FA- or CAP-exposed mice were euthanized on the next day after the last exposure.

CAP Exposure

Animal exposure and the monitoring of exposure atmosphere and ambient aerosol were performed in the animal facility of UMB with a 12-h light/12-h dark cycle, temperatures of 24°–26°C, and relative humidity of 40%–60%, as previously described (Qiu et al. 2018). The exposure protocol comprised exposures for 6 h/d, 5 d/wk (no exposure during weekends) for 16 wk. Briefly, the CAP was generated through the concentration of ambient using the versatile aerosol concentration enrichment system (VACES) in the animal facility of UMB, and the FA was generated with the addition of a Whatman® in-line high-efficiency particulate air (HEPA) filtration unit (WHA67235000, Sigma) in the inlet valve to remove CAP in the stream. CAP generated using this VACES and ambient were shown to have comparable elemental compositions (Chen et al. 2018; Ying et al. 2014). To assess the mass concentration of , ambient and in the FA and CAP streams were collected onto Teflon™ filters [Teflo, 37-mm, pore (Pall Life Sciences)]. The mass of on a filter would be the difference in its weights in a temperature- and humidity-controlled room using a Mettler Toledo no. 11106057 microbalance (Mettler Toledo) before and after sampling. The mass concentrations of for each experiment were calculated using the sampling flow volumes and presented in Table 1 and the figure legends. To determine the composition of , the CAP samples collected on Teflon filters were wetted with ethanol and extracted in 1% nitric acid solution for 2 wk following a 48-h sonication in an ultrasonic bath. A suite of trace elements in these extracts were analyzed using inductively coupled plasma–mass spectrometry (ICP-MS) (ELEMENT2; Thermo Finnigan). Briefly, the samples were introduced through a standard Meinhard nebulizer with a cyclonic spray chamber. Instrumental parameters were previously described (Morishita et al. 2004). Repeated samples and instrument-induced variations were compensated using internal standardization. Reagent and filter blanks were also analyzed for background elemental content, and appropriate corrections were made. The composition of CAP in the present study is presented in Table 2.

Table 1

The concentrations of ambient and in the FA or CAP exposure chamber.

	Ambient	FA	CAP
Exposure 1	11.3±4.3	2.8±0.9	79.3±61.9
Exposure 2	12.8±5.7	3.1±1.3	82.2±67.4

Note: All data are presented as (). , 4, and 16 for Ambient, FA, and CAP, respectively. Exposure 1, exposure of and control mice; Exposure 2, exposure of and control mice. CAP, concentrated ambient ; FA, filtered air; PM, particulate matter; POMC, pro-opiomelanocortin; SD, standard deviation.

Table 2

The compositions of CAP.

Element	Exposure 1		Exposure 2
	Mean	SD	Mean	SD
Na	5.0959	2.8507	4.0414	1.0189
Mg	1.2507	0.6976	2.4301	0.4203
Al	3.7597	1.8367	4.8448	2.4065
Si	11.3885	5.7125	16.3683	9.3813
P	0.1449	0.0722	0.1699	0.0719
S	36.2413	26.5995	19.7255	5.7577
K	2.9549	1.6778	6.0778	1.8230
Ca	5.6987	2.8802	9.5700	3.5044
Ti	0.5153	0.3033	1.4994	0.6096
V	0.0380	0.0216	0.2991	0.0701
Cr	0.0407	0.0224	0.3521	0.1350
Mn	0.1729	0.1090	1.0616	0.3421
Fe	6.8429	4.3656	11.5349	2.9203
Ni	0.0936	0.0573	0.9275	0.2759
Cu	0.2109	0.1342	1.9919	0.3799
Zn	1.9976	0.9890	3.2561	1.1024
As	0.0274	0.0140	0.1051	0.0559
Se	0.0188	0.0094	0.1187	0.0801
Br	1.2217	0.7569	1.2409	0.3930
Rb	ND	0.0000	0.0711	0.0420
Sr	0.0341	0.0187	0.6281	0.0648
Ag	0.0802	0.0277	0.0476	0.0413
Sn	0.1813	0.0776	0.0615	0.0403
Ba	0.2415	0.1682	0.5197	0.1193
Ce	0.0354	0.0359	ND	0.0000
Pr	0.0429	0.0187	ND	0.0000
Er	0.1428	0.1083	0.4778	0.1620
Lu	0.1930	0.1722	0.8204	0.2186
W	0.0593	0.0381	0.1266	0.0678
Ir	0.0157	0.0063	ND	0.0000
Pt	0.0285	0.0135	ND	0.0000
Au	0.0312	0.0205	ND	0.0000
Hg	0.0494	0.0074	0.0868	0.0507
Tl	0.0157	0.0056	ND	0.0000
Pb	0.1347	0.0466	0.2907	0.1002
Mo	ND	0.0000	0.1224	0.0747
Zr	ND	0.0000	0.0902	0.0475
Cd	ND	0.0000	0.0420	0.0097

Note: All data are presented as (). . Exposure 1, exposure of and control mice; Exposure 2, exposure of and control mice. CAP, concentrated ambient ; FA, filtered air; ND, not detectable; PM, particulate matter; POMC, pro-opiomelanocortin; SD, standard deviation.

Analysis of Epididymal Sperm

Freshly harvested epididymides of each mouse were put in normal saline, cut at the epididymis cauda section with surgical scissors, and incubated at 35°C for 10 min. The released sperm were then filtered out using nylon mesh with a pore size of (Sigma) and stained with 0.25% eosin Y containing 0.5% formalin (Sigma). The number of sperm was determined using a Neubauer counting chamber as previously described (Qiu et al. 2018).

Testicular Histological Analysis

The right testis of each mouse was fixed in the Bouin’s solution for 24 h, embedded in paraffin, and cut into sections. A total of 9 sections per mouse were then subjected to hematoxylin and eosin (H&E) staining. The images of these sections were obtained using Olympus IX70 microscope and the cellSens Standard software (Olympus). ImageJ was used for the quantitation analysis. Briefly, all Stage VII tubules (identified by the presence of preleptotene spermatocytes, per section) were identified as previously described (Ahmed and de Rooij 2009). Their diameters, wall heights, and numbers of Sertoli cells and germ cells including spermatogonia, pachytene spermatocytes, and round spermatids were assessed. The numbers of Sertoli cells and germ cells were then corrected by the nuclear diameter of each cell type and the section thickness as previously described (George et al. 1996). The abundance of these cells was expressed as the corrected number per seminiferous tubule.

Hormone Assessments

Each mouse was euthanized using intraperitoneal injection of 200 mg/kg sodium pentobarbital euthanasia solution. The blood was collected from the trunk after decapitation and clotted at room temperature for 30 min. The serum was harvested after centrifuging at for 10 min in a refrigerated centrifuge, aliquoted, and kept at . Enzyme-linked immunosorbent assay kits were used to determine the levels of serum and testicular hormones including follicle-stimulating hormone (FSH; LH-E10075MU; LIUHEBIO), testosterone (T; IB79174; IBL-American), and estradiol (E2; ES180S-100; Calbiotech) per the manufacturer’s instruction.

Immunohistochemistry Analysis

This analysis was conducted as previously described (Qiu et al. 2016). Briefly, six sections of each mouse described above were immersed in citrate buffer (, pH 6.0) and heated in a GE 1.1-cubic foot mid-size microwave for 5 min for antigen retrieval, blocked in 1% BSA for 2 h at room temperature, incubated with 1:100 diluted anti-IKK beta antibody (EPR6043) (ab124957) overnight at 4°C. After washing, sections were incubated with 1:1,000 diluted goat anti-rabbit IgG () cross-adsorbed secondary antibody, HRP (G-21,234; Thermo Fisher Scientific), and visualized with the 3,3′ diaminobenzidine (DAB) enzyme substrate color development kit (TA-060-QHDX; Thermo Fisher Scientific) per the manufacturer’s instruction. The images were captured using an Olympus IX70 system and cellSens software.

Quantitative Polymerase Chain Reaction (qPCR)

RNAs (, determined using a NanoDrop™ 2000 from Thermo Fisher Scientific) of the left testis and hypothalamus of each mouse were isolated using Trizol reagent (Invitrogen) and reverse-transcribed to cDNA using High Capacity cDNA Reverse Transcription Kits (Thermo Fisher Scientific, Inc.) per the manufacturer’s instructions. PCR was performed using a LightCycler® 480 (Roche), cDNA, and 2× FastStart™ Universal SYBR® Green Master system (Roche). The sequences of primers used in this study are presented in Table 3. The specificity of each PCR reaction was confirmed by melting curve analysis. The relative expression of every gene vs. glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was calculated as and presented as the fold to the control group.

Table 3

The sequences of primers.

Primer	Sequence
AR	Sense: 5′-CTG GGA AGG GTC TAC CCA C-3′
	Antisense: 5′-GGT GCT ATG TTA GCG GCC TC-3′
P450scc	Sense: 5′-AGG TCC TTC AAT GAG ATC CCT T-3′
	Antisense: 5′-TCC CTG TAA ATG GGG CCA TAC-3′
P450arom	Sense: 5′-AGG TCC TTC AAT GAG ATC CCT T-3′
	Antisense: 5′-AGG ACC TGG TAT TGA AGA CGA G-3′
CYP17α	Sense: 5′-GCC CAA GTC AAA GAC ACC TAA T-3′
	Antisense: 5′-GTA CCC AGG CGA AGA GAA TAG A-3′
HSD17β	Sense: 5′-ACT TGG CTG TTC GCC TAG C-3′
	Antisense: 5′-GAG GGC ATC CTT GAG TCC TG-3′
HSD3β	Sense: 5′-CCT CCG CCT TGA TAC CAG C-3′
	Antisense: 5′-TTG TTT CCA ATC TCC CTG TGC-3′
StAR	Sense: 5′-ATG TTC CTC GCT ACG TTC AAG-3′
	Antisense: 5′-CCC AGT GCT CTC CAG TTG AG-3′
SHGB	Sense: 5′-TCT GCT GTT GCT ACT ACT GAT GC-3′
	Antisense: 5′-GGG CCA TTG CTG AGG TAC TTA-3′
GnRH	Sense: 5′-AGC ACT GGT CCT ATG GGT TG-3′
	Antisense: 5′-GGG GTT CTG CCA TTT GAT CCA-3′
IL-6	Sense: 5′-ATC CAG TTG CCT TCT TGG GAC TGA-3′
	Antisense: 5′-GGG GTT CTG CCA TTT GAT CCA-3′
IL-1β	Sense: 5′-ACG GAC CCC AAA AGA TGA AG-3′
	Antisense: 5′-TTC TCC ACA GCC ACA ATG AG-3′
GAPDH	Sense: 5′-TGA ACG GGA AGC TCA CTG G-3′
	Antisense: 5′-TCC ACC ACC CTG TTG CTG TA-3′

Note: AR, androgen receptor; , cytochrome P450 family ; GAPDH, glyceraldehydes-3-phosphate dehydrogenase; GnRH, gonadotropin-releasing hormone; , dehydrogenase; , dehydrogenase; IL-6, interleukin-6; , ; P450arom, cytochrome P450 aromatase; P450scc, cytochrome P450 CHOL side-chain cleavage enzyme; SHGB, sex hormone binding globulin; StAR, steroidogenic acute regulatory protein.

Statistics

All data are expressed as unless noted otherwise. Statistical tests were performed using one-way or two-way analysis of variance (ANOVA) followed by Bonferroni correction or unpaired -test (specified in the figure legend of each analysis) using GraphPad Prism (version 8.0.2; GraphPad Software, Inc.). The significance level was set at .

Results

Measurements of Hypothalamic Inflammation, the HPG Axis, and Spermatogenesis of Mice following CAP Exposure

CAP-exposed mice vs. FA-exposed mice had significantly higher expressions of hypothalamic and mRNAs. In contrast, CAP-exposed mice and FA-exposed mice had comparable expressions of hypothalamic , IL-6, and mRNAs (Figures 1A–C). Next, we assessed the effects of CAP exposure on the HPG axis in both and littermate mice. In comparison with FA-exposed mice, CAP-exposed mice had a significantly lower expression of hypothalamic GnRH mRNA, significantly lower levels of circulating and testicular FSH and testosterone (Figures 1D and 1H–K), but comparable levels of circulating and testicular estrogen (Figures 1L and 1M). CAP-exposed mice vs. FA-exposed mice also had significantly lower expressions of testicular target genes of testosterone, including StAR, P450scc, , and SHBG (Figure S2). In contrast, CAP-exposed mice and FA-exposed mice had comparable functional indexes of the HPG axis (Figures 1D and 1H–K).

Figure 1.

The effects of neural deficiency of IKK2 on CAP exposure-induced abnormalities in the HPG axis and spermatogenesis. and littermate mice (8–11 wk old, ) were exposed to FA (Mean concentration: ) or CAP (Mean concentration: ) for 16 wk. (A–D) The expression of indicated mRNAs relative to GAPDH mRNA in the hypothalamus determined by qPCR. (E–G) The testicular expression of indicated mRNAs relative to GAPDH mRNA determined by qPCR. (H–M) The levels of indicated testicular or circulating hormones determined by ELISA. (N) The counts of epididymal sperm. (O) The weights of their freshly isolated testis, normalized by their BW. (P) The weights of their freshly isolated epididymis, normalized by their BW. (Q) Their percentages of the morphologically abnormal epididymal sperms. (R) The mean diameters of seminiferous tubules at Stage VII. (S) The mean wall heights of seminiferous tubules at Stage VII. (T–X) The abundances of indicated germ cells in the Stage VII tubule and (T–W) the representative histological images (X). All the data were presented as . Black arrowhead: immature germ cells. Bar in the upper panel: . Bar in the lower panel: . * vs. FA, two-way ANOVA. Summary data for all graphs is presented in Excel Table S1. mRNA data plotted relative to FA-exposed for all genotypes and exposures. Note: ANOVA, analysis of variance; BW, body weight; CAP, concentrated ambient ; ELISA, enzyme-linked immunosorbent assay; FA, filtered air; GADPH, glyceraldehyde-3-phosphate dehydrogenase; HPG, hypothalamic–pituitary–gonadal; qPCR, quantitative real-time polymerase chain reaction; SD, standard deviation.

We next tested whether exposure impairs sperm production through a hypothalamic inflammation-dependent mechanism. We found that CAP-exposed mice vs. FA-exposed mice had a significantly lower epididymal sperm count (Figure 1N), comparable morphologically abnormal sperm, and comparable weights of testis and epididymis (Figures 1O–Q). Sperm are produced in the testicular seminiferous tubules. Notably, the seminiferous tubules of CAP-exposed mice vs. those of FA-exposed mice had comparable diameters and wall heights (Figures 1R and 1S). The histological abnormality of CAP-exposed mice vs. FA-exposed mice was accompanied by significantly less pachytene spermatocytes (Figures 1V and 1X) and round spermatids (Figures 1W and 1X) but comparable spermatogonia (Figures 1U and 1X) and Sertoli cells (Figures 1T and 1X) in their seminiferous tubules. In contrast, CAP-exposed mice vs. FA-exposed mice were comparable in all these measurements (Figures 1R–X).

Given the pivotal role of IKK2 in inflammatory responses (Arkan et al. 2005), it was necessary to rule out the possibility that testicular IKK2 was unexpectedly deleted in mice. We thus assessed testicular expression of IKK2 in both and control mice through immunostaining. We found that neural IKK2 deficient mice and littermate controls had comparable expression levels of testicular IKK2 protein (Figure S3). Furthermore, CAP-exposed mice vs. FA-exposed mice had a comparable expression of testicular proinflammatory cytokine mRNAs, including , , and IL-6 mRNAs (Figures 1E–G).

Measurements of Hypothalamic Inflammation, the HPG Axis, and Spermatogenesis of Mice

We generated a neuron-specific IKK2ca overexpression mouse model (the mice). The adult mice and littermate mice were macroscopically indistinguishable (Figure S1B). In comparison with littermate controls, the adult mice had a significantly higher expression of , , and IL-6 mRNAs in the hypothalamus (Figures 2A–C). In line with the specificity of this Nestin-Cre line, the expression of these inflammatory markers in the testis of mice was comparable to that of littermate controls (Figures 2E–G). In contrast, mice vs. littermate controls had a significantly lower expression of hypothalamic GnRH mRNA (Figure 2D), significantly lower levels of circulating and testicular testosterone (Figures 2H and 2I) but not estradiol (Figures 2L and 2M), and a significantly lower expression of testicular target genes of testosterone, including StAR, , and (Figures S4A, E, and F, respectively) but not P450scc, P450arom, , AR, and SHBG (Figures S4B, C, D, G, and H, respectively). We then counted the epididymal sperm of and littermate control mice. mice vs. controls had a significantly lower epididymal sperm count but comparable abnormal epididymal sperm and weights of testis and epididymis (Figures 2N-2Q). In addition, mice vs. controls had comparable diameters and wall heights (Figures 2R and 2S) but less pachytene spermatocytes (Figures 2V and 2X) and round spermatids (Figures 2W and 2X) in the wall of seminiferous tubules. In contrast, their spermatogonia (Figures 2U and 2X) and Sertoli cells (Figures 2T and 2X) in the wall of seminiferous tubules were comparable.

Figure 2.

The effects of neural overexpression of IKK2ca on the HPG axis and spermatogenesis. and littermate mice (20–26 wk old, ) were euthanized and assessed for hypothalamic inflammation and spermatogenesis. (A–D) The expression of indicated mRNAs relative to GAPDH mRNA in the hypothalamus determined by qPCR. (E–G) The testicular expression of indicated mRNAs relative to GAPDH mRNA determined by qPCR. (H–M) The levels of indicated testicular or circulating hormones determined by ELISA. (N) The counts of epididymal sperm. (O) The weights of their freshly isolated testis, normalized by their BW. (P) The weights of their freshly isolated epididymis, normalized by their BW. (Q) Their percentages of the morphologically abnormal epididymal sperms. (R) The mean diameters of seminiferous tubules at Stage VII. (S) The mean wall heights of seminiferous tubules at Stage VII. (T–X) The abundances of indicated germ cells in the Stage VII tubule (T–W) and the representative histological images (X). All the data were presented as . Black arrowhead: immature germ cells. Bar in the upper panel: . Bar in the lower panel: . * vs. , Student -test. Summary data for all graphs is presented in Excel Table S1. mRNA data plotted relative to . Note: BW, body weight; ELISA, enzyme-linked immunosorbent assay; GADPH, glyceraldehyde-3-phosphate dehydrogenase; HPG, hypothalamic–pituitary–gonadal; qPCR, quantitative real-time polymerase chain reaction; SD, standard deviation.

Measurements of Hypothalamic Inflammation, the HPG Axis, and Spermatogenesis of and Mice

POMC neurons and AgRP neurons are two main types of hypothalamic neurons. We overexpressed IKK2ca specifically in POMC neurons ( mice) or AgRP neurons ( mice). mice vs. littermate controls had a significantly higher expression of hypothalamic and IL-6 mRNAs and mice vs. littermate controls had a significantly higher expression of hypothalamic IL-6 mRNA (Figures 3A–C). In contrast, neither nor mice vs. littermate controls had a significant difference from their counterparts in the testicular expression of these inflammatory markers (Figures 3E–G). Notably, the differences between mice and littermate controls in the HPG axis (Figures 3D and 3H–3M), epididymal sperm count (Figure 3N), and spermatogenesis (Figures 3O–X and S5) mimicked almost all the differences between CAP-exposed mice and FA-exposed mice (Figure 1). In contrast, no significant difference in these measurements was observed between mice and littermate controls (Figure 3).

Figure 3.

The effects of hypothalamic neuron type-specific overexpression of IKK2ca on the HPG axis and spermatogenesis. , , and respective littermate control mice (18–23 wk old, ) were euthanized and assessed for hypothalamic inflammation and spermatogenesis. (A–D) The expression of indicated mRNAs relative to GAPDH mRNA in the hypothalamus determined by qPCR. (E–G) The testicular expression of indicated mRNAs relative to GAPDH mRNA determined by qPCR. (H–M) The levels of indicated testicular or circulating hormones determined by ELISA. (N) The counts of epididymal sperm. (O) The weights of their freshly isolated testis, normalized by their BW. (P) The weights of their freshly isolated epididymis, normalized by their BW. (Q) Their percentages of the morphologically abnormal epididymal sperms. (R) The mean diameters of seminiferous tubules at Stage VII. (S) The mean wall heights of seminiferous tubules at Stage VII. (T–X) The abundances of indicated germ cells in the Stage VII tubule (T–W) and the representative histological images (X). All the data were presented as . Black arrowhead: immature germ cells. Bar in the upper panel: . Bar in the lower panel: . * vs. littermate controls, student t test. Summary data for all graphs is presented in Excel Table S1. mRNA data plotted relative to respective genotype. Note: BW, body weight; ELISA, enzyme-linked immunosorbent assay; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HPG, hypothalamic–pituitary–gonadal; qPCR, quantitative real-time polymerase chain reaction; SD, standard deviation.

Given that overexpression of IKK2ca in POMC neurons, but not in AgRP neurons, mimicked the effects of exposure on the HPG axis and spermatogenesis, we generated mice deficient of IKK2 in POMC neurons ( mice) and tested whether this deficiency of IKK2 is sufficient to block the effects of exposure on the HPG axis and spermatogenesis. FA-exposed mice vs. FA-exposed littermate controls had a comparable expression of testicular IKK2 protein (Figure S6). Although CAP-exposed mice vs. FA-exposed mice had a significantly higher expression of hypothalamic and IL-6 mRNAs, CAP-exposed mice vs. FA-exposed mice had comparable levels of hypothalamic inflammatory markers (Figures 4A–C). In contrast, neither CAP exposure nor this genetic manipulation of IKK2 expression significantly influenced the expression of testicular inflammatory markers (Figures 4E–G). Furthermore, although CAP-exposed mice vs. FA-exposed mice had a significantly lower expression of hypothalamic GnRH mRNA (Figure 4D), significantly lower levels of circulating and testicular FSH and testosterone but not estradiol (Figures 4H–M), a significantly lower expression of testicular P450scc and mRNAs (Figure S7), a significantly lower epididymal sperm count (Figure 4N) but not testicular weight, epididymal weight, and abnormal sperms (Figures 4O–Q), and significantly less pachytene spermatocytes and round spermatids in the seminiferous tubules (Figures 4V–X), CAP-exposed mice vs. FA-exposed mice were comparable in all these measurements.

Figure 4.

The effects of POMC neuron-specific deficiency of IKK2 on CAP exposure-induced abnormalities in the HPG axis and spermatogenesis. and littermate mice (8-10 wk old, ) were exposed to FA (Mean concentration: ) or CAP (Mean concentration: ) for 16 wk and assessed for hypothalamic inflammation and spermatogenesis. (A–D) The expression of indicated mRNAs relative to GAPDH mRNA in the hypothalamus determined by qPCR. (E–G) The testicular expression of indicated mRNAs relative to GAPDH mRNA determined by qPCR. (H–M) The levels of indicated testicular or circulating hormones determined by ELISA. (N) The counts of epididymal sperm. (O) The weights of their freshly isolated testis, normalized by their BW. (P) The weights of their freshly isolated epididymis, normalized by their BW. (Q) Their percentages of the morphologically abnormal epididymal sperms. (R) The mean diameters of seminiferous tubules at Stage VII. (S) The mean wall heights of seminiferous tubules at Stage VII. (T–X) The abundances of indicated germ cells in the Stage VII tubule (T–W) and the representative histological images (X). All data are presented as . Black arrowhead: immature germ cells. Bar in the upper panel: . Bar in the lower panel: . * vs. FA, two-way ANOVA. Summary data for all graphs is presented in Excel Table S1. mRNA data plotted relative to FA-exposed for all genotypes and exposures. Note: BW, body weight; CAP, concentrated ambient ; ELISA, enzyme-linked immunosorbent assay; FA, filtered air; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HPG, hypothalamic–pituitary–gonadal; POMC, pro-opiomelanocortin; qPCR, quantitative real-time polymerase chain reaction; SD, standard deviation.

Discussion

Considerable studies, both in humans (Carré et al. 2017; Lafuente et al. 2016) and animal models (Cao et al. 2015; Qiu et al. 2018; Wei et al. 2018; Yang et al. 2019; Zhou et al. 2019), have demonstrated that exposure reduces sperm production and thus may be one of the primary culprits for the continuing worldwide decline in the sperm count. However, the underlying mechanism for this impairment of sperm production by exposure is not yet fully understood. Hypothalamic inflammation now emerges as a crucial mediator for the pathophysiology due to exposure (Chen et al. 2018; Liu et al. 2014; Qiu et al. 2018; Sun et al. 2018; Ying et al. 2014). The present study, therefore, thoroughly tested the role of hypothalamic inflammation in the impairment of sperm production by exposure. The major findings of this study are that a) deficiency of neural IKK2 blocked both hypothalamic inflammation and disruptions of the HPG axis and sperm production due to CAP exposure; b) like CAP exposure, neural overexpression of IKK2ca elicited remarkable hypothalamic inflammation and disruptions of the HPG axis and sperm production; c) overexpression of IKK2ca in POMC neurons but not in AgRP neurons was sufficient to elicit the disruptions of the HPG axis and sperm production; and d) deficiency of IKK2 in POMC neurons blocked both hypothalamic inflammation and disruptions of the HPG axis and sperm production induced by CAP exposure. All these findings strongly support the hypothesis of a causal role of hypothalamic inflammation in the impairment of sperm production by exposure, thereby providing a deep mechanistic insight into this mounting public health concern induced by exposure.

Hypothalamic inflammation, typically manifested as the activation of IKK2/NF-KB signaling pathway and expression of proinflammatory cytokines in the hypothalamus, is an emerging mediator for various occurrences of pathogenesis (Lainez and Coss 2019b). Several studies demonstrate that exposure results in hypothalamic inflammation (Campolim et al. 2020; Dong et al. 2021; Qiu et al. 2018; Xu et al. 2020; Ying et al. 2014). Consistent with these published studies, this study revealed that exposure resulted in a significantly higher expression of hypothalamic proinflammatory cytokines. A more important finding is that the present study clearly showed that blocking exposure-induced expression of hypothalamic proinflammatory cytokines through deleting neural IKK2 halted the impairment of sperm production induced by exposure. To the best of our knowledge, this finding is the strongest current evidence for the essential role played by hypothalamic inflammation in the impairment of sperm production by exposure. Furthermore, this study also shows that hypothalamic inflammation subsequent to overexpression of IKK2ca in POMC neurons was sufficient to mimic the impairment of sperm production by exposure. These data collectively demonstrate a clear covariation of hypothalamic inflammation and impairment of sperm production in the response to exposure, indicating a causal role of hypothalamic inflammation in the impairment of sperm production by exposure.

Another important finding of this study is that in the pathophysiology associated with exposure, hypothalamic inflammation also covaried with disruption of the HPG axis, unmasking a cascade that may link hypothalamic inflammation to the impairment of sperm production. This covariation is evidenced by our data showing that deficiency of neural IKK2 blocked both hypothalamic inflammation and disruption of the HPG axis induced by CAP exposure and that hypothalamic inflammation subsequent to neural overexpression of IKK2ca provoked disruption of the HPG axis. These data are consistent with published studies showing that inflammation down-regulates the hypothalamic expression of GnRH mRNA (George et al. 2010; Kalra et al. 1998; Purkayastha and Cai 2013). Notably, a previous study suggested that inflammation might directly inhibit the testicular production of testosterone (Foster et al. 2003), and several other studies showed that exposure provoked testicular inflammation (Cao et al. 2015; Wei et al. 2018; Zhou et al. 2019), offering a hypothalamic inflammation–independent mechanism for exposure to affect the HPG axis and thus the sperm count. However, in our mouse models, this study provides at least three lines of evidence ruling out any important role of this mechanism in exposure impacting the HPG axis: a) CAP exposure did not elicit a significant expression of testicular inflammatory markers; b) CAP exposure decreased the expression of GnRH and FSH, in direct contrast to their supposed increase due to the feedback from the inhibition of testicular production of testosterone; and c) blocking hypothalamic inflammation through deficiency of neural IKK2 restored CAP exposure-induced disruption of the HPG axis. The reason for this discrepancy between the present study and those published studies remains to be determined.

By far, although considerable studies have shown that hypothalamic inflammation affected the HPG axis and thus the reproductive system in various inflammation models (Lainez and Coss 2019a; Wu and Wolfe 2012), the underlying mechanism remains elusive. The present study demonstrated in mice that the POMC neurons but not the AgRP neurons mediated the disruption of the HPG axis and thus the sperm production by the hypothalamic inflammation due to exposure, providing a deep insight into the mechanism for the action of hypothalamic inflammation on the HPG axis and subsequently sperm production. Specifically, our demonstration of the POMC neuron-dependency for hypothalamic inflammation affecting the HPG axis strongly suggests that synaptic connections but not paracrine or endocrine signaling is involved in this process, as the apposition of POMC neurons and AgRP neurons in the hypothalamus are well known (Rønnekleiv et al. 2019). This implication of synaptic connections in the disruption of the HPG axis by hypothalamic inflammation is consistent with previous studies showing that many proinflammatory mediators may affect hypothalamic neural activity (Peng et al. 2011; Reis et al. 2015) and that there are synaptic connections between POMC neurons and GnRH neurons (or kisspeptin neurons, the command neurons driving pulsatile release of GnRH) (Rønnekleiv et al. 2019). Furthermore, the effect of hypothalamic inflammation on the HPG axis through synaptic connections is also strongly supported by mounting studies suggesting that exposure provokes hypothalamic inflammation through afferent vagal signaling (Dalby et al. 2018; De Souza et al. 2005; Gaykema et al. 2008; Marvel et al. 2004; Naznin et al. 2018; Waise et al. 2015) but not circulating proinflammatory mediators (Chen et al. 2017; Hu et al. 2017; Ying et al. 2015).

In conclusion, the present study demonstrated a crucial role of hypothalamic IKK2/NF-KB signaling in the impairment of sperm production by exposure in mice, strongly supporting the causal role of hypothalamic inflammation in this pathophysiology.

Click here for additional data file.

Click here for additional data file.

Click here for additional data file.