All procedures involving live animals in this study were reviewed and approved by the Laboratory Animal Management and Use Ethics Committee of Shenyang Agricultural University (Approval No. 20240513). The experiment was conducted in full compliance with the institutional guidelines for the care and use of experimental animals and the relevant national regulations on animal welfare in China (including the Regulations on the Administration of Laboratory Animals and the Guidelines for the Ethical Review of Laboratory Animal Welfare).

The study was performed at the LCG Breeding Center, Liaoyang City, Liaoning Province, China. All blood sample collections were carried out under the direct supervision of a licensed veterinarian to ensure minimal stress and discomfort to the animals. No invasive surgical procedures were performed, and all handling was conducted according to the principles of the 3Rs (Replacement, Reduction, and Refinement). The animals remained under standard farm management conditions throughout the study, and no additional pain, suffering, or distress beyond routine husbandry practices was imposed.

Written informed consent for the use of the animals was obtained from the farm management prior to the commencement of the study. The authors confirm that the research adheres to the ARRIVE 2.0 guidelines (Animal Research: Reporting of In Vivo Experiments) for the reporting of animal studies.

This study was conducted from April to August 2025 at the LCG Breeding Center, Liaoyang City, Liaoning Province, China (approximate coordinates: 41.24° N, 123.14° E).

A total of 1,160 clinically healthy LCG were enrolled in the study (89 bucks and 1,071 does; age range 2–4 years). All animals belonged to the same nucleus breeding herd and were therefore at a comparable production stage. Inclusion criteria were as follows: clinically healthy appearance, absence of major parasitic infestations, and no history of antibiotic treatment within the preceding 4 weeks. Animals were maintained under semi-intensive management conditions with a stocking density of approximately 15–20 goats per pen. They had ad libitum access to a total mixed ration and clean drinking water. To maximize genetic diversity within the sampled population, individuals were deliberately selected from multiple unrelated families.

All phenotypic measurements were performed by the same fixed team of five trained technicians to ensure maximum consistency and to minimize inter-observer variation. Detailed definitions, measurement timing, units, and protocols for each recorded trait are presented in Supplementary Table S1. The main instruments and corresponding calibration procedures are described below.

CF (fiber diameter) and staple length were determined using the portable all-weather CF and length analyzer (Model: OFDA2000, BSC Electronics, Ardross, Australia). The instrument was calibrated weekly before each measurement session using a certified standard cashmere bundle (nominal fineness 15.0 ± 0.1 μm) according to the Chinese national standard Test Method for Cashmere Fiber (GB/T 18267-2019). The calibration deviation was maintained at ≤ 0.2 μm throughout the study. Collected cashmere samples were evenly spread on a dedicated sample slide, the slide was inserted into the detection chamber, and the “Start Test” function was activated to record fiber diameter and length.

Body conformation measurements were obtained using an intelligent three-dimensional optical body measurement system (3D Optical Body Scanner, Model: VITUS Smart, Human Solutions GmbH, Kaiserslautern, Germany). Animals were gently guided into the measurement zone where the system automatically initiated multi-angle scanning. Recorded parameters included: body height, withers height, body length, chest depth, chest width, hip width, heart girth, cannon circumference, and pin bone height. The instrument was calibrated monthly using a standard reference mannequin, achieving a measurement error of ≤ 0.5 cm.

Milk composition was analyzed using the MilkoScan FT120 milk analyzer (Foss Electric A/S, Hillerød, Denmark). The following parameters were determined: milk fat, milk protein, lactose, urea nitrogen, non-fat solids, total solids, conductivity, and health index. Daily calibration was performed before each analysis session using Foss calibration milk (Product No.: 8011-0001, Foss Electric) according to ISO 9622:2013. Calibration error for the major components (fat, protein, lactose) was maintained at ≤ 0.05%. The built-in conductivity sensor was simultaneously calibrated (measurement range 0–20 mS/cm; accuracy ±0.05 mS/cm).

Whole blood was collected from the jugular vein into commercial ethylenediaminetetraacetic acid vacuum blood collection tubes (Sangon Biotech, Shanghai, China). Immediately after collection, tubes were placed on ice, transported to the laboratory within 4 hours, and stored at –20 °C until DNA extraction (maximum storage duration 2 weeks).

Genomic DNA was extracted from 200 µL of anticoagulated whole blood using the Ezup column animal genomic DNA extraction kit (Cat. No. B518253-0100, Sangon Biotech, Shanghai, China) following the manufacturer’s protocol. In brief: Proteinase K and lysis buffer (Buffer DL) were added, followed by incubation at 56°C for 10 min. After addition of absolute ethanol and thorough mixing, the lysate was transferred to the adsorption column. The column was washed sequentially with GW solution and Wash solution (two washes each), followed by a high-speed centrifugation step to remove residual Wash buffer. DNA was finally eluted with Capillary Electrophoresis Buffer. DNA purity and concentration were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Samples were considered acceptable when the A260/A280 ratio ranged between 1.8 and 2.0 and the concentration exceeded 50 ng/µL. Qualified DNA samples were stored at –20°C until further use.

The target SNP loci C31799T (GMPS) and C34197G (SRC) were initially discovered and confirmed by bidirectional Sanger sequencing of PCR amplicons obtained from a pilot subset of 30 animals (15 bucks and 15 does). Obtained sequences were aligned against the goat reference genome (GenBank accession nos. NC_030808.1 and NC_030820.1) using DNAMAN software version 9.0. Only variants confirmed by both sequencing directions with a minimum Phred quality score of Q30 were retained.

Gene-specific primers flanking the target intronic regions of GMPS and SRC were designed using Primer Premier 5.0 based on the goat reference genome (NC_030808.1 and NC_030820.1). All primer pairs were designed to have GC content of 45%–55% and were screened against potential primer-dimer formation. Primer sequences, optimal annealing temperatures (Tm), and expected amplicon sizes are presented in Table 1.

PCR was performed in a total volume of 50 µL containing:

25 µL of 2× SanTaq PCR Mix (Sangon Biotech, Shanghai, China)

Amplification products were separated on 1.5% (w/v) agarose gels prepared in 1× TAE buffer and stained with GelRed nucleic acid stain. Electrophoresis was conducted at 130 V (constant power 180 W) for 20 min. A 100 bp DNA ladder (Thermo Fisher Scientific) was included for size determination. Amplicons showing a single band of the expected size were purified using the SanPrep Column PCR Product Purification Kit (Sangon Biotech, Shanghai, China) according to the manufacturer’s instructions. Purified products were submitted for bidirectional Sanger sequencing (Sangon Biotech Co., Ltd., Shanghai, China).

Genotypes were manually called from bidirectional sequencing chromatograms using Chromas software version 2.6.6. A heterozygous call was accepted only when the secondary peak height reached at least 35% of the primary peak height. All chromatograms were independently scored by two researchers. Samples exhibiting poor sequence quality or a genotype call missing rate >5% were excluded. To assess repeatability, the complete workflow (PCR → sequencing → genotyping) was repeated for 10% of randomly selected samples, resulting in 100% genotype concordance.

Haplotype phase was inferred using the SHEsis online platform with default parameters. To allow detection of sex-specific haplotype effects, haplotype reconstruction was performed separately for bucks and does. Rare haplotype combinations (population frequency <1%) were collapsed into a single “other” category for subsequent association analyses.

Figure 1 shows the agarose gel electrophoresis results of PCR products for the target loci. (Left panel) A single specific band of 754 bp corresponding to the GMPS amplicon. (Right panel) A single specific band of 637 bp corresponding to the SRC amplicon. Successful and specific amplification suitable for downstream Sanger sequencing was confirmed by comparison with a 100 bp DNA ladder.

Figure 2 presents representative bidirectional sequencing chromatograms of the target polymorphic sites. (Left) GMPS C31799T locus showing homozygous CC, heterozygous CT, and homozygous TT genotypes. (Right) SRC C34197G locus showing homozygous CC, heterozygous CG, and homozygous GG genotypes. Arrows indicate the polymorphic nucleotide positions; clear double peaks in heterozygous samples confirm reliable detection of heterozygosity.

Genotype frequencies, allele frequencies, polymorphism information content (PIC), expected heterozygosity (He), effective number of alleles (Ne), and Hardy–Weinberg equilibrium χ² test results at the two loci are presented in Table 2. The T allele at GMPS C31799T was predominant in both sexes (frequency 0.81 in bucks, 0.78 in does). The C allele at SRC C34197G was predominant (frequency 0.80 in bucks, 0.81 in does). PIC values ranged from 0.26 to 0.28, indicating moderate polymorphism at both loci. Significant deviation from the Hardy–Weinberg equilibrium was observed in most sex–locus combinations (p < 0.01).

Additive and dominance effects estimated for the two loci are summarized in Table 3. Substitution of the C allele by the T allele at GMPS C31799T produced positive effects on production performance (average substitution effect 45.33% in bucks, 233.72% in does). Substitution of the C allele by the G allele at SRC C34197G produced negative effects (average substitution effect –25.11% in bucks, –386.07% in does).

Genotype means ± standard error and significance levels for cashmere production traits are presented in Table 4. In bucks, the GMPS TT genotype was associated with significantly longer staple length compared with CC (p < 0.01, partial η² = 0.096). In does, the GMPS TT genotype conferred superior CF (p < 0.01, partial η² = 0.025) and longer staple length (p < 0.01, partial η² = 0.022) compared with CC. The SRC CG genotype in does was associated with significantly finer cashmere fibers compared with CC (p < 0.01, partial η² = 0.045).

Different lowercase letters indicate significant differences (p < 0.05); different uppercase letters indicate highly significant differences (p < 0.01). Sample sizes were limited for some genotype groups (e.g., GMPS CC in bucks n = 6, GMPS CC in does n = 14).

Effects of GMPS and SRC on body size traits in LCG Genotype means ± standard error for body conformation traits are shown in Table 5. The GMPS CC genotype was associated with superior body size traits in both sexes. In bucks, CC conferred significantly greater height at the sacrum and waist height (p < 0.01, partial η² = 0.319 and 0.216, respectively). In does, CC was linked to significantly greater sacrum height, chest depth, chest width, and waist width (p < 0.01, partial η² = 0.009–0.020). The SRC CG genotype in does was associated with significantly greater body height, chest depth, and waist height compared with CC and GG (p < 0.05, partial η² = 0.012–0.067).

Different lowercase letters indicate significant differences (p < 0.05); different uppercase letters indicate highly significant differences (p < 0.01). Sample sizes were limited for some genotype groups (e.g., GMPS CC in bucks n = 6).

Genotype means ± standard error for slaughter performance traits are presented in Table 6. In bucks, the GMPS TT genotype was superior for live weight before slaughter, net meat percentage of carcass, and back fat thickness (p < 0.01, partial η² = 0.151–0.305). In does, TT was associated with higher slaughter rate and GR value (p < 0.01, partial η² = 0.542–0.947). For SRC in does, the CC genotype showed advantages in live weight before slaughter, carcass weight, net meat weight, and GR value (p < 0.01, partial η² = 0.225–0.419).

Different lowercase letters indicate significant differences (p < 0.05); different uppercase letters indicate highly significant differences (p < 0.01). Sample sizes were limited for some genotype groups (e.g., GMPS CC in bucks n = 3, SRC GG in bucks n = 10).

Genotype means ± standard error for meat quality parameters are presented in Table 7. In bucks, the GMPS CC genotype was associated with superior meat color a* value and dry matter content (p < 0.01, partial η² = 0.296 and 0.499, respectively). The SRC CG genotype in bucks showed more favorable meat color L* and b* values and higher protein content compared with CC (p < 0.01, partial η² = 0.772–0.991). In does, the SRC CC genotype demonstrated advantages in protein content, fat content, and shear force compared with CG (p < 0.01, partial η² = 0.259–0.420).

Milk composition parameters by genotype are shown in Table 8. For GMPS in does, the CT and TT genotypes exhibited superior performance compared with CC for fat, crude protein, urea nitrogen, solids-not-fat, and total solids (p < 0.05, partial η² = 0.010–0.029). For SRC, the CC and CG genotypes demonstrated better performance than GG for fat, urea nitrogen, solids-not-fat, and total solids (p < 0.01, partial η² = 0.040–0.064).

Litter size means by genotype are presented in Table 9. No significant differences in the number of lambs per doe were detected across genotypes at either the GMPS C31799T or SRC C34197G loci.

Pearson correlation coefficients between CF and other fiber and production traits are shown in Table 10. Fiber fineness exhibited highly significant positive correlations with shearing quantity (bucks r = 0.41, does r = 0.38; p < 0.01) and highly significant negative correlations with cashmere yield rate (bucks r = –0.90, does r = –0.93; p < 0.01) in both sexes.

Direct and indirect path coefficients from production traits to CF are shown in Table 11. In does, the strongest direct negative effect was exerted by cashmere yield rate (–0.92), followed by coefficient of variation of length (0.10). In bucks, cashmere yield rate (–0.89) and shearing quantity (0.23) were the dominant direct contributors.

Multiple linear regression models with CF as the dependent variable are presented in Table 12. For bucks: CF = 0.001 × shearing quantity – 0.07 × cashmere yield rate + 20.553 (R² = 0.85). For does: CF = 0.0001 × shearing quantity – 0.079 × cashmere yield rate + 22.307 (R² = 0.86). Both models explained more than 85% of the variation in CF.

Pearson correlation coefficients between CF and slaughter traits are shown in Table 13. Fineness exhibited highly significant negative correlations with several carcass merit indicators, including slaughter rate, net meat consumption rate, and net carcass meat percentage in both sexes (p < 0.01).

Direct and indirect path coefficients from slaughter traits to CF are presented in Table 14. In bucks, net meat weight exerted the strongest direct positive effect on fineness (31.82). In does, the slaughter rate was the primary direct negative contributor (–1.05).

Optimal multiple regression equations for CF based on slaughter traits are shown in Table 15. For does: CF = 27.558 – 0.217 × slaughter rate (R² = 0.23). For bucks: CF = 36.723 – 0.230 × net carcass meat percentage – 0.118 × eye muscle area (R² = 0.55).

Nine haplotype combinations were inferred from the two SNP loci using SHEsis software (Table 16). Bucks displayed six combinations: CCCC, CCCG, CTCC, CTCG, TTCC, TTCG. Does displayed seven combinations: CCGG, CTCC, CTCG, CTGG, TTCC, TTCG, TTGG.

Haplotype means ± standard error for cashmere traits are presented in Table 17. In bucks, the CCCC combination was associated with the finest cashmere fibers. In does, the TTGG combination outperformed the other haplotypes in fineness, coefficient of variation of length, and curl number.

Haplotype means ± standard error for slaughter performance traits are presented in Table 18. In bucks, TTCC was the superior haplotype for live weight before slaughter, carcass weight, and back fat thickness. In does, CTCC demonstrated consistent superiority across multiple slaughter traits.

This deviation is likely attributable to multiple factors. Primarily, the long-term, intensive artificial selection practiced to improve CF and growth performance has likely led to directed changes in allele frequencies and the evolution of the population’s genetic structure [17]. Additionally, factors such as the limited size of the breeding population, skewed sex ratios, or a certain degree of inbreeding may have exacerbated this deviation [18]. In our statistical analyses, we accounted for these potential confounding factors by incorporating pedigree information (to control for kinship), and by including sex and age as fixed effects in a general linear model. This approach was implemented to mitigate the interference of these factors on the detected association signals.

GMPS and SRC are evolutionarily conserved genes with pivotal biological functions across species. GMPS functions as a key rate-limiting enzyme in the purine synthesis pathway, responsible for catalyzing the production of guanosine monophosphate (GMP), thereby supplying essential substrates for cellular proliferation. Given the high metabolic demand during hair follicle growth and cyclic regeneration, we speculate that GMPS may influence CF by modulating the proliferative activity of follicular keratinocytes. Although initial functional characterization was primarily conducted in model organisms such as bacteria, yeast, and insects [19–21], emerging evidence reveals complex regulatory capacities beyond canonical enzymatic activity, and subsequent studies have substantiated its significant role in tumor biology. Notably, GMPS has been identified as a critical molecular hub that integrates intracellular mechanisms with extracellular microenvironmental signals across various cancers. Building upon these findings, we propose that GMPS may regulate CF in LCGs through a dual mechanism: (1) as a metabolic enzyme providing purine nucleotides for rapidly dividing follicle cells during fiber synthesis, and (2) as a signaling modulator potentially involved in follicular microenvironment communication, analogous to its role in tumor-stroma interactions.

On the other hand, the SRC family exerts central regulatory functions in core physiological processes, including reproduction, metabolism, and the stress response, by significantly enhancing the transcriptional activity of various steroid hormone receptors [22]. It is crucial to distinguish this SRC coactivator from the SFKs, which are non-receptor tyrosine kinases involved in intracellular signal transduction [4]; this study focuses on the former. Steroid hormones (e.g., estrogen, progesterone) are key regulators of hair follicle development and cycle. We hypothesize that SRC, as a universal transcriptional coactivator, may amplify localized hormonal signaling within skin follicles, thereby modulating the proliferation and differentiation of follicular stem cells, ultimately influencing secondary follicle density and CF. Substantial evidence supports its vital role in development; for instance, SRC1 and SRC2 act synergistically to regulate extracellular matrix remodeling and multi-organ development, which is essential for ensuring normal fetal lung maturation, a critical signal event for initiating parturition. Furthermore, SRC is highly likely to mediate progesterone-dependent transcription of genes like Crispld2 during lung development [23]. By analogy, SRC may orchestrate gene expression related to cashmere follicle growth and fiber synthesis in a similar manner. Consequently, polymorphisms in SRC might concurrently affect cashmere traits and slaughter performance by modulating hormone responsiveness and associated metabolic pathways. This viewpoint finds indirect support in ovine studies confirming that SRC is an essential molecular switch that enables glucocorticoids to exert their organ-protective effects, thereby precisely coordinating hormonal signaling and gene expression [24, 25].

GMPS and SRC are hub genes regulating critical life processes, with well-defined functions in human medicine and fundamental physiology. This study focused on their roles in LCGs. Through SNP genotyping and association analysis, we found that polymorphisms at the GMPS C31799T and SRC C34197G loci not only significantly influenced CF but also exhibited pleiotropic effects on other production traits. The superior genotypes displayed distinct and sex-dependent patterns that may reflect a resource-allocation trade-off between fiber synthesis and muscle growth, coordinated by pleiotropic genes. For instance, the T allele of the GMPS gene might preferentially direct metabolic resources toward cashmere production, resulting in finer fibers, whereas the C allele appears more beneficial for body growth and muscle deposition. Similarly, the C allele of the SRC gene seems to contribute positively to both CF and meat quality. Path analysis further elucidated these complex relationships. In bucks, net meat weight exhibited a significant positive direct effect on fiber diameter (direct path coefficient: 31.82), while net meat rate showed the strongest negative direct effect (–29.92), revealing a genetic antagonism between meat and cashmere traits. In does, both slaughter rate and carcass net meat percentage demonstrated consistent negative direct effects (–1.05 and –0.52, respectively), and their indirect effect pathways were entirely different from those in bucks. This supports the argument for sex-specific genetic regulation underlying the correlation of these traits. In summary, these results suggest that GMPS and SRC may co-regulate key economic traits such as CF, body size, and slaughter performance by coordinating resource allocation between follicular development and muscle metabolism, within a sex-specific genetic regulatory architecture.

The effects of GMPS and SRC genes identified in the LCG in this study show both commonalities and distinctions compared to findings in other cashmere goat breeds. Population genetics studies indicate genetic differentiation between LCG and some other local Chinese breeds [26], suggesting that the genetic architecture underlying CF may exhibit breed specificity. The key candidate genes influencing fineness appear to differ across breeds. For instance, in Shaanbei White Cashmere goats, polymorphisms in genes such as Notch2 have been significantly associated with cashmere growth and body size traits [27]. Conversely, in Inner Mongolia Cashmere goats, the primary genetic determinants of fineness may involve different loci. Polymorphisms in the prolactin receptor (PRLR) gene have been linked to CF [28], while variations in the keratin-associated protein KAP8.1 gene predominantly affect cashmere yield without a significant impact on fineness [29]. These findings collectively suggest that the key genes regulating CF in Inner Mongolia Cashmere goats are likely distinct from the GMPS and SRC genes identified in our study. Furthermore, recent genomic analyses of Tibetan Northwest White Cashmere goats reveal that their finer cashmere fibers are associated with specific genomic regions under strong selection, which differ significantly from those in plain-breed cashmere goats [30]. These comparisons underscore the diverse genetic foundations of key economic traits in cashmere goats. Our study is the first to report significant associations between GMPS/SRC polymorphisms and CF in LCG, thereby contributing novel molecular markers to the expanding toolkit for molecular breeding in cashmere goats.

The present study demonstrates a significant correlation between cashmere yield and cashmere length, with the greatest indirect effect coefficient being 0.318. Through research conducted by Feng et al. [31] on Shaanbei White Cashmere goats, it was found that the correlation coefficient between cashmere yield and cashmere length was the highest, followed by chest circumference, body weight, and cannon circumference. Studies by He Xiaohong et al. [32] indicated that cashmere length had the greatest impact on cashmere yield, chest circumference had the strongest influence on body weight, while body slope length, cannon circumference, and staple length indirectly affected body weight through chest circumference. Research by Li Longping et al. [33] revealed extremely significant positive correlations among cashmere yield, cashmere length, CF, and body weight in Shaanbei White Cashmere goats, with interactions observed among these traits (cashmere length, fineness, body weight, etc.). Based on the aforementioned literature, we hypothesized whether traits correlated with CF exist within both cashmere production and slaughter performance. Our correlation analysis between CF and slaughter traits revealed significant correlations between CF and slaughter rate, net meat rate, net meat percentage of carcass, and eye muscle area. Furthermore, correlation analysis between CF and various traits showed significant correlations between CF and cashmere yield, cashmere length, coefficient of variation of length, number of crimps, short fiber content, and cashmere rate. These findings are consistent with reports by Shi Lei et al. [34], who found a highly significant positive correlation between CF and length, and Zhang Fuquan et al. [35], who reported a positive correlation between CF and yield. This suggests that the physiological processes that control fiber fineness may be genetically or metabolically coupled to muscle development, although the specific mechanisms of interaction require further investigation.

This study reveals statistically significant associations rather than direct causal relationships. Nonetheless, the identified superior genotypes and haplotype combinations provide valuable pre-selection markers for molecular breeding in LCGs. The observed negative correlation between CF and slaughter traits may stem from pleiotropy, tight genetic linkage, or indirect effects of physiological resource allocation, constituting a key focus for subsequent functional studies to elucidate the specific molecular interactions. In practical breeding, where the core objectives are to consistently reduce fiber diameter, each micrometer reduction can significantly enhance market value, while maintaining high yield, we propose a MAS strategy involving: 1) the establishment of sex-specific selection protocols, and 2) the development of criteria for early genetic evaluation and retention of replacement breeding stock based on the significant effects of GMPS and SRC. Validation of these markers in independent populations is expected to improve selection accuracy for fineness, shorten the genetic improvement cycle, and thus enable a steady reduction in fiber diameter while preserving cashmere yield. Future work should prioritize validation in larger populations to confirm stability, followed by research to identify key causal mutations and clarify the expression mechanisms of these genes within hair follicles, thereby strengthening the theoretical foundation and technical support for future breeding programs.