Healthy American blood donors (n = 392) were recruited at the Memorial Blood Center (737 Pelham Boulevard, St. Paul, Minnesota 55114) as previously described (29). The age of healthy blood donors ranged from 19 to 84 years-old with the mean age of 64 ± 13.5 years and >98% of donors in the study were self-declared Caucasians living in the State of Minnesota. Sarcoidosis patients (n = 59) for CD64 expression phenotype analysis were recruited at the University of Minnesota Medical Center Interstitial Lung Disease Clinic. The ages of sarcoidosis patients ranged from 28.7 to 79.2 years-old with the mean age of 55.4 ± 13.7 years. In addition, human samples and data from the ACCESS (A Case Control Etiologic Study of Sarcoidosis) cohort were used in the genetic association analysis. The DNA specimens and data of the ACCESS sarcoidosis patients (670 cases) and healthy controls (669 controls) were provided by the NHLBI Biologic Specimen and Data Repository as previously described (28). The human study was approved by the Institutional Review Board for Human Use at the University of Minnesota (IRB Protocol #1301M26461).

The expressions of CD64 on monocytes were determined with flow cytometry analysis with fluorescent-labeled monoclonal antibodies from BioLegend (San Diego, CA). Leukocytes stained with either FITC-conjugated anti-CD64 mAb (10.1, mIgG1) plus APC-conjugated anti-CD14 or mIgG1-FITC isotype control plus APC-conjugated anti-CD14 were analyzed on a FACS Canto flow cytometer (BD Biosciences). The FlowJo software (Tree Star Inc.) was used to evaluate flow cytometry data. Characteristic light-scatter properties were used to identify neutrophils and CD14⁺ monocytes cells were gated in flow cytometry for the geometric mean intensity of CD64 expression.

Human genomic DNA was isolated from EDTA anti-coagulated peripheral blood using the Wizard Genomic DNA Purification kit (Promega, Madison, WI) by following the vendor’s instruction.

LaserGene DNAstar MegAlign software (Madison, WI) was used to identify the FCGR1A gene-specific sequence regions within the FCGR1A promoter and 3’-UTR through alignment of nucleotide sequences of FCGR1A (RefSeq AL591493), FCGR1B (RefSeq AL357493.8), and FCGR1C (RefSeq AL109948.9). The long-template PCR with the sense primer (5’-TTA GCT CTC TTT AGC TCT CTT TTT TTA GCT CTC AT -3’) and antisense primer (5’- CCT CGG TAG GTC CCA GGG AGA AGA AAG ATT C -3’) was carried out to produce the FCGR1A gene-specific DNA fragment (11,685 bps) containing the proximal promoter region and all six exons of FCGR1A ( Figure 1 ). PCR was performed using 200 ng genomic DNA, 200 nM of each primer, 200 μM of dNTPs, 1× PrimeStar GXL buffer, and 2 U of PrimeStar GXL DNA polymerase (TaKaRa cat# R050), starting with 98°C for 1 min; 30 cycles of denaturing at 98°C for 10 s, annealing and extension at 68°C for 5 min; with a final extension at 68°C for 7 min on an ABI Veriti 96-well Thermal Cycler. PCR products were treated with ExoSAP-IT (Affymetrix, Santa Clara, CA) before being sequenced on an ABI 3730xl DNA Analyzer with BigDye Terminator Cycle Sequencing Kit. Five sequencing primers were used for sequence analyses: 1) 5’-CAG CCA CAG CCT GTA CCC TT-3’ for the proximal promoter region and exon 1 coding for the signal peptide, 2) 5’-CAC AAA AGC CTC ACC AGT TGC-3’ for the extracellular domain 1 (EC1), 3) 5’-CTT GGG CCT CCT TGT ACC TCC-3’ for the extracellular domain 2 (EC2), 4) 5’-GGT AAA GGG CAT GTC TTT TGT GA-3’ for the extracellular domain 3 (EC3), and 5’-ATG TTT GTA CGC AGT GCT CA-3’ for the transmembrane segment and intracellular domain (TMC). Nucleotide sequence tracers were aligned and compared with the FCGR1A genomic sequence (RefSeq NG_007578.1) using the DNASTAR software (DNAStar, Madison, WI) for the identification of FCGR1A variants. All variations are described according to current mutation nomenclature guidelines (30), assigning the A of the first ATG translational initiation codon as nucleotide +1 in the FCGR1A mRNA coding region (RefSeq NM_000566.3). Genomic DNA samples of 102 human subjects from the cohort of 392 healthy blood donors were used for FCGR1A sequence analysis.

The FCGR1A promoter reporter constructs were generated by cloning a Kpn I/Bgl II-flanked FCGR1A promoter DNA (819 bps) fragment into pGL4.23[luc2/minP] vector (Cat# E841A, Promega, Madison, WI). The Kpn I/Bgl II-flanked DNA products were generated by PCR amplification of the genomic DNA from a SNV rs1848781 (c.-131C>G) heterozygous donor using upper primer 5’-CGG GGT ACC TCT TTA GCT CTC TTT TTT TAG CTC TCA-3’ (underlined and bold nucleotides are Kpn I cutting site, the primer anneals at position from -794 to -831) and lower primer 5’-CGC AGA TCT GTT GTC TCC AAG CTG GTG GG-3’ (underlined and bold nucleotides are Bgl II cutting site, the primer anneals at position from -20 to -1). The nucleotide sequences of the cloned constructs were confirmed by direct sequencing from both directions on an ABI 377 Sequencer with ABI BigDye Terminator Cycle Sequencing Kit.

Human monocytic cell line U-937 (ATCC# CRL-1593.2) was obtained from ATCC (Manassas, VA). The cells were maintained in the RPMI-1640 medium supplemented with 10% fetal calf serum and L-glutamine (2 mM). The transient transfections were carried out in a 12-well tissue culture plate (Corning). The cells (2 × 10⁵ cells per well) in 2 ml culture medium were transiently transfected with 3 μl of TransIT-2020 reagent (Mirus Bio LLC, Madicon, WI), 1 μg reporter construct plasmid DNA, and 0.025 μg pRL-SV40 plasmid DNA (Promega, Madison, WI) by following vendor’s instruction. The transfected cells were cultured for 20 hours in the presence of IFN-γ (100 U/ml) or absence of IFN-γ before being centrifuged and washed twice with PBS (pH 7.4). The cells were lysed in the wells with the addition of 200 μl of 1× lysis buffer for the Luciferase Assay Systems (Promega, Madison, WI). The cell supernatants were used for luciferase reporter assays by following vendor’s instruction. Relative luciferase light units, standardized to Renilla luciferase activities in dual luciferase reporter assays, are reported as the mean of triplicate samples.

The human FcγRIA (CD64) expression constructs were generated by cloning EcoR I/Bam HI-flanked RT-PCR products with the entire FcγRIA coding region into the cDNA Cloning and Expression Lentivector pCDH-MSCV-MCS-IRES-copGFP (System Biosciences, Mountain View, CA). The Kpn I/Bam HI-flanked FcγRIA cDNA was amplified from human mixed mononuclear cell cDNA synthesized with the SuperScript^TM Preamplification System (Gibco BRL) with the sense primer (5’-CGC GAA TTC GGA GAC AAC ATG TGG TTC TTG ACA A-3’) (underlined and bold nucleotides are EcoR I cutting site) and anti-sense primer (5’-CCG GGA TCC CTA CGT GGG CCC CTG GGG CTC CTT -3) (the underlined and bold nucleotides is Bam HI restriction cutting site). The PCR reaction was performed with 2 μl of cDNA, 200 nM of each primer, 200 μM of dNTPs, 2.0 mM of MgSO₄, and 1 U of Platinum Taq DNA polymerase High Fidelity (Invitrogen) in a 25 μl reaction volume. The ABI Veriti 96-well Thermal Cycler was used for the PCR reaction starting with 94°C for 3 min, 35 cycles of denaturing at 94°C for 30 s, annealing at 56°C for 45 s, extension at 68°C for 1 min and 30 s with a final extension at 72°C for 7 min. The QuikChange Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA) and two mutagenesis primers [the sense primer 5’-AAG AAA AAG TGG A AT TTA GAA ATC TC-3’ and anti-sense primer 5’-GAG ATT TCT AAA T TC CAC TTT TTC TT-3’ (underlined and italicized letter is the intentional mutation)] were used to change the amino acid residue from 324D to 324N in the FcγRIA expression construct. The sequences of all the cloned constructs were confirmed by direct sequencing from both directions on an ABI 377 Sequencer with ABI BigDye Terminator Cycle Sequencing Kit.

Pseudo-viral particles of pCDH-MSCV-MCS-IRES-copGFP expression constructs were produced with the packaging plasmids according to the vendor’s manual (System Biosciences). The murine macrophage cell line P388D1 (P388) and rat mast cell line RBL-2H3 (RBL) obtained from ATCC (Manassas, VA) were maintained in the DMEM medium supplemented with 10% fetal calf serum (FBS) and L-glutamine (2 mM). Transductions of lentiviral particles were carried out on a 60-mm cell culture dish with the cell density at ~80% confluence. Pseudo-viral particles in 5 ml culture medium were mixed with polybrene (final concentration: 4 µg/ml) before being added to the target cell dishes. The medium was removed, and fresh medium was added one day after transduction. Two days after transduction, the CD64 expression was detected by FACS analyses. The transduced cells were sorted on a FACSAria II (BD Biosciences) for equal expression of all constructs in both the P388D1 and RBL-2H3 cells.

P388 cells expressing equivalent levels of either FcγRIA-p.324D (rs1050204G) or FcγRIA-p.324N (rs1050204A) were used to determine the phagocytosis capacity of two FcγRIA alleles. An adherent phagocytosis assay with the probe consisting of biotinylated bovine erythrocytes (aka EB) and biotinylated anti-CD64 mAb 32.2 F(ab′)₂ was used to examine phagocytosis capacity of FcγRIA alleles in P388D1 cells as previously described (6, 31, 32). EB were saturated with streptavidin to form erythrocyte-biotin-avidin (aka EBA). After a wash step, the EBA were coated with biotinylated anti-CD64 mAb 32.2 F(ab’)₂ (aka EBA-32.2) and the levels of mAb 32.2 binding to EBA were verified by flow cytometry. P388D1 cells adhered to round glass coverslips were incubated with EBA-32.3 in medium for 1 hour at 37 °C. Non-internalized bovine erythrocytes were lysed by brief immersion of the coverslip in distilled H₂O followed by immersion in buffer. Three negative controls were included (1): the parental P388D1 cells with EBA-32.2 (2); the P388 cells expressing FcγRIA-p.324D with EBA, and (3) the P388 cells expressing FcγRIA-p.324N with EBA. Phagocytosis was quantitated using light microscopy and presented as the phagocytic index (number of bovine erythrocytes internalized per 100 P388D1 cells). At least 200 cells per slide were counted in duplicate without knowledge of the cell types.

RBL cells (10⁵ cells/well) expressing either FcγRIA-p.324D (rs1050204G) or FcγRIA-p.324N (rs1050204A) allele were cultured overnight in 24-well culture plates (Corning). The culture media were removed from plates and the cells were incubated with DMEM medium containing either anti human CD64 mAb F(ab’)₂ (clone 32.2, final concentration of 5 µg/ml) or mouse IgG (mIgG) F(ab’)₂ for 45 min at 4°C. The cells were washed with Tyrode’s buffer (130 mM NaCl, 5 mM KCl, 1.4 mM CaCl₂, 1 mM MgCl₂, 5.6 mM glucose, 10 mM HEPES, and 0.1% BSA, pH 7.4) and then stimulated with goat anti mouse F(ab’)₂ (Jackson ImmunoResearch, West Grove, PA) at final concentration of 20 μg/ml in the Tyrode’s buffer. The supernatants were collected at 0, 15, 30, 45, and 60 min for measurement of β-hexosaminidase activity. Cells from control wells were lyzed with the same volume of 0.1% Triton X-100 in Tyrode’s buffer for evaluation of the total β-hexosaminidase activity for each RBL stable cell lines. Supernatants and cell lysates were incubated with substrate (1.3 mg/ml p-nitrophenyl-N-acetyl β-D-glucosamine) (Sigma, St. Louis, MO) in 0.1 mM sodium citrate (pH 4.5) for 1 hour at 37°C. The reaction was stopped by 0.2 mM sodium carbonate buffer (pH 10.0) and the enzyme reactivity was evaluated by measuring optical density at 405 nm. The percentage of specific β-hexosaminidase activity released was calculated as follows: percentage release (%) = 100 × average supernatant activity from 4 wells/average cell lysate activity from 4 control wells.

P388 cells were stimulated in 24-well tissue culture plates (Corning) with surface-bound anit-CD64 mAb 32.2 F(ab’)₂ as previously described (33). Wells were coated with either anti-CD64 mAb 32.2 F(ab’)₂ (20 μg/ml) or control mIgG F(ab’)₂ (20 μg/ml) overnight at room temperature. The CD64-expressing cells were added to coated wells and the culture media were collected after 24 hours. The levels of murine Il-1β, IL-6, or TNFα in culture media were quantified by BD™ Cytometric Bead Array (CBA) kits (cat# 558279 for murine Il-1β, cat# 558301for IL-6, and cat# 558299 for TNFα respectively, BD Biosciences).

High-throughput TaqMan genotyping assays of FCGR1A variants were used to determine genotypes of human subjects. Two FCGR1A gene-specific DNA fragments (the 1,009-bps fragment containing the SNV rs1848781 and the 2,131-bps fragment containing the indel variant rs587598788 and the SNV rs1050204) were amplified by PCR using genomic DNA and primers listed on Table 1 . Standard TaqMan reactions were subsequently performed with 1 µl of the respective FCGR1A gene-specific PCR products, the primers, and fluorescence-labeled (FAM or Vic) labeled probes ( Table 1 ). The FCGR1A genotypes were determined using Applied Biosystems 7500 Software. The TaqMan assay genotypes were compared to those determined by Sanger sequencing methodology. A perfect (100%) concordance of genotypes between TaqMan assay and direct sequencing analysis was achieved in all 102 human subjects, confirming the specificity and accuracy of FCGR1A TaqMan genotyping assays.

To determine associations between individual FCGR1A variants and sarcoidosis susceptibility, additive logistic regressions were performed either with or without race, age, and sex as covariates for each FCGR1A variant. Odds ratios with 95% confidence intervals were computed using the profile likelihood method of The R Project for Statistical Computing (version 4.1.2, https://www.R-project.org/). To account for the multiple testing corrections, the FDR-corrected P-values were generated by using False Discovery Rate (FDR) adjustment. For haplotype analyses, additive haplotype models were fit using logistic method, either without or with covariates. Odds ratios, p-values, and 95% confidence intervals were reported. All calculations performed using the haplo.stats R package version 1.7.9. (haplo.stats: Statistical Analysis of Haplotypes with Traits and Covariates when Linkage Phase is Ambiguous. https://CRAN.R-project.org/package=haplo.stats).

The rst_pfts (restriction on pulmonary function tests) positive (rst_pfts+) patients had worse or poorer lung functions than rst_pfts negative (rst_pfts-) patients. To determine associations between individual FCGR1A variant and sarcoidosis lung functions, we divided sarcoidosis patients into two groups based on rst_pfts status. Additive logistic regressions were performed for each FCGR1A variant with rst_pfts status (positive or negative) as response variables. Odds ratios with 95% confidence intervals were computed using the profile likelihood method of The R Project for Statistical Computing.

Unpaired t-tests (Mann-Whitney test) were used to analyze CD64 expression levels on monocytes and neutrophils between sarcoidosis patients and healthy controls and among healthy donors stratified with FCGR1A variant genotypes. The Student’s t-test was used to analyze the data for promoter reporter, cytokine production, and degranulation assays. A P-value less than 0.05 was considered as significant.

FCGR1A gene contains six exons. The exon 1 and 2 (S1 and S2) code for the CD64 signal peptide while the exon 3, 4, and 5 code for the extracellular domain 1 (EC1), 2 (EC2), and 3 (EC3), respectively. The CD64 transmembrane segment and cytoplasmic domain (TMC) are coded by the exon 6 ( Figure 1 ). The full-length FCGR1A gene fragment (11,685 bps) containing proximal promoter and all six exons was amplified with a long-template PCR and subsequently analyzed by direct Sanger sequencing. By sequencing the FCGR1A promoter and exons of 102 human subjects, we did not detect any non-synonymous SNVs within FCGR1A exons coding for three CD64 extracellular domains responsible for the interaction with IgG ligands. On the other hand, as shown in Figure 1 , a common SNV at the nucleotide position -131 (rs1848781 or c.-131C>G) was identified in the FCGR1A proximal promoter region. In addition, a sole non-synonymous SNV (rs1050204 or c.970G>A) that changes the amino acid codon position 324 (p.D324N) from aspartate (D) to asparagine (N) in the FcγRIA cytoplasmic domain was identified in the exon 6. Furthermore, we detected a novel indel variant (rs587598788 or c.845-23_845-17delTCTTTG) within intron 5 that causes six-nucleotide insertion or deletion near the splicing acceptor site of the exon 6. The genotype distributions of those three variants were consistent with the Hardy-Weinberg equilibrium in 102 healthy blood donors (P > 0.05).

To examine relationship between FCGR1A variant genotypes and CD64 expressions, we carried out genotype-phenotype analyses in healthy blood donors. In a discovery cohort, we found that the FCGR1A SNV rs1848781 (c.-131C>G) genotypes were significantly associated with CD64 expressions on resting monocytes of healthy blood donors ( Figure 2 ). Monocytes from C/G (-131C/G) heterozygous donors expressed significantly higher levels of CD64 than those from the C/C (-131C/C) homozygous donors (P = 0.0077). Monocytes from two G/G (-131G/G) homozygous donors had the highest average levels of CD64 among three genotype groups ( Figure 2 ). The significant association of FCGR1A SNV rs1848781 (c.-131C>G) genotypes with CD64 expression levels was confirmed by a replication cohort consisting of 275 human subjects (P < 0.0001, Figure 2 ). We also carried out promoter reporter assays to determine whether the SNP -131C>G influences the promoter activity. As shown in Figure 2 , the promoter reporter construct containing the SNV rs1848781G (or -131G) allele had significantly higher promoter activity than that with -131C allele in human monocytic U937 cells in the presence of IFN-γ, consistent with the observation that monocytes from the donors carrying -131G allele (-131C/G heterozygous and -131G/G homozygous donor) expressed higher levels of CD64 than those from the -131C/C homozygous donors. Taken together, data from ex vivo monocytes and in vitro promoter reporter assays demonstrate that the SNV rs1848781 (c.-131C>G) is a functional SNV affecting the gene expression.

We further analyzed effects of FCGR1A rs587598788 and rs1050204 variants on CD64 expressions. As shown in Figure 3 , resting monocytes from the rs587598788-c.845-23 (rs587598788In) homozygous donors (In/In) expressed significantly higher levels of CD64 than the combined group of rs587598788 heterozygous (In/Del) and rs587598788-c.845-17delTCTTTG (rs587598788Del) homozygous donors (Del/Del) (P = 0.0136). The significant association between CD64 expression levels and rs587598788 genotypes was confirmed in a replication cohort (P = 0.0009, Figure 3 ). On the other hand, the SNV rs1050204 genotypes were not associated with CD64 expression (P = 0.4888, Figure 3 ), which was consistent with the equivalent or similar expression of CD64 in the cell lines expressing rs1050204G (FcγRIA-p.324D) and rs1050204A (FcγRIA-p.324N) alleles ( Figures 4A, B ).

Previous studies demonstrated that FcγRIA cytoplasmic domain influences receptor functions (6, 32, 34). The FCGR1A SNV rs1050204 (c.970G>A or p.D324N) leads to non-conservative residue change from p.324D to p.324N in the FcγRIA cytoplasmic domain that participates in signal transduction. Consequently, we investigated whether the FCGR1A SNV rs1050204 (FcγRIA-p.D324N) alleles affect FcγRIA-mediated functions using P388D1 cell lines expressing equivalent levels of FcγRIA-p.324D (rs1050204G) and FcγRIA-p.324N (rs1050204A). Phagocytosis assay was carried out as described in “Materials and Methods” and all negative controls showed no phagocytosis of bovine erythrocytes. FcγRIA-p.324N mediated significantly higher levels of phagocytosis (phagocytosis index = 41.00 ± 3.46) than the FcγRIA-p.324D (phagocytosis index = 25.75 ± 1.03) (P = 0.0056) ( Figure 4 ). In addition, the FcγRIA-p.324N allele mediated significantly more degranulation than the FcγRIA-p.324D allele ( Figure 4 ) and significantly more pro-inflammatory cytokine (IL-6, IL-1β, and TNFα) productions than did the FcγRIA-p.324D allele ( Figure 5 ). Our data demonstrate that the FCGR1A SNV rs1050204 (FcγRIA-p.D324N) within CD64 cytoplasmic domain significantly affect receptor-mediated functions, similar to the effect of the SNV within FcαRI cytoplasmic domain (33).

We subsequently carried out a genetic study to examine whether FCGR1A variants are associated with sarcoidosis susceptibility using the ACCESS cohort subjects. As previously described, the major goal of ACCESS was to address the hypotheses that sarcoidosis occurs in genetically susceptible individuals through alteration in immune response after exposure to an environmental, occupational, or infectious agents (35, 36). Specific phenotypes of sarcoidosis were determined with an instrument developed by the ACCESS group (37). The clinical characteristics of the study patients have been described previously (38). As shown in Table 2 , FCGR1A SNV rs1020204 genotypes were associated with sarcoidosis susceptibility (P = 0.042, P_FDR = 0.111, OR 1.222, 95% CI 1.007 – 1.546) while rs587598788 genotypes tended to associate with sarcoidosis (P = 0.073, P_FDR = 0.111, OR 1.431, 95% CI 0.962 – 2.128). On the other hand, SNV rs1848781 genotypes were not associated with sarcoidosis risk (P = 0.747) ( Table 2 ).

FCGR1A variants in linkage disequilibrium could form different haplotypes to impact gene functions. Subsequently, we carried out haplotype analysis to examine whether FCGR1A variant haplotypes are associated with the sarcoidosis susceptibility. As shown in Table 3 , we found that the haplotype C-Del-A (rs1848781C-rs587598788Del-rs1050204A) was significantly associated with the protection against sarcoidosis development (logistic regression adjusted for sex and age, P = 0.008, OR 0.542, 95% CI 0.346 – 0.853). The frequency of the haplotype C-Del-G (rs1848781C-rs587598788Del-rs1050204G) containing the rs1848781C-rs587598788Del also tended to be lower in sarcoidosis patients (1.95%) than that in the matched controls (2.4%). Taken together, our data suggest that FCGR1A haplotypes may affect the pathogenesis of sarcoidosis.

Sarcoidosis frequently affects lung functions (14). We analyzed whether FCGR1A variants are associated lung functions among sarcoidosis patients. The restriction on pulmonary function tests (rst_pfts) status was used to stratify sarcoidosis patients as described in “Materials and Methods”. As shown in Table 4 , FCGR1A rs1848781 genotypes were significantly associated with the risk for poor lung function (genotype C/G+G/G vs C/C, P = 0.001, P_FDR = 0.016, OR 1.695, 95% CI 1.263 – 2.262). In addition, the rs1050204 genotypes tended to associate with the risk for poor lung function as well (genotype A/G+A/A vs G/G, P = 0.035, P_FDR = 0.263, OR 1.471, 95% CI 1.028 – 2.079). On the other hand, no significant difference in the distribution of rs587598788 genotypes was observed between rst_pfts positive and rst_pfts negative sarcoidosis patients, possibly due to the low frequency of the rs587598788Del allele in populations. Our data suggest FCGR1A variant genotypes are risk factors for poor lung function in sarcoidosis patients, pointing to a role of FCGR1A in lung inflammation.

To investigate a role of FcγRIA in sarcoidosis, we examined expression levels of CD64 in sarcoidosis patients and healthy blood donors. We found that CD64 expressions on monocytes ( Figures 6A, B ) and neutrophils ( Figures 6C, D ) were significantly increased in sarcoidosis patients (SA) as compared to healthy controls (HC) (P < 0.001). Our data indicate that the upregulation of CD64 could be a biomarker for sarcoidosis.