Analysis of maternal microchimerism in rhesus monkeys (Macaca mulatta) using real-time quantitative PCR amplification of MHC polymorphisms

Analysis of maternal microchimerism levels in rhesus monkeys using Mamu-MHC allele-specific qPCR assays

Maternal microchimerism testing was initially performed on 25 monkeys. Six offspring were excluded from analysis because maternal allele-specific qPCR assays could not be identified. Microchimerism analysis was performed on DNA extracted from tissues (thymus, liver, spleen, lymph nodes, and bone marrow) collected prenatally from seven fetuses and postnatally from 12 juveniles. In the latter group, testing was also done on DNA extracted from peripheral blood mononuclear cells (PBMCs) collected both prenatally and postnatally from four of 12 fetuses, whereas for the remaining eight neonates, testing was done on DNA extracted from PBMCs collected only at or after birth.

Each maternal-fetal pair was first genotyped using a panel of Mamu-MHC allele-specific qPCR assays (see Materials and Methods) to identify assays informative for investigating maternal microchimerism. Subsequently, the fetal DNA sample (containing at least 10⁵ genomic equivalents of input DNA) and three 10-fold serial dilutions of maternal DNA were amplified in parallel using the assays for the informative MHC alleles and for GAPDH. Melting curve analysis and gel electrophoresis were used to verify that any signal detected was specific in both fetal and maternal DNA samples, and not due to nonspecific products or primer dimer artifacts. Results for a representative assay from one pair are shown in Figure 1A. The GAPDH assay Ct for each maternal sample dilution was plotted against the MHC informative allele assay Ct. A standard curve was generated and analyzed by performing linear regression. For the example shown in Figure 1B, the frequency of the maternal allele within the fetal sample, calculated using the ΔCt method, was 2^-10.6 or 0.06%. To further verify specificity of the signal detected in the fetal sample, the PCR products amplified from the maternal and fetal DNA samples were purified and sequenced. The sequence obtained from the fetal sample was identical to that from the maternal sample (Figure 1C).

Detection of maternal microchimerism in fetal rhesus monkey tissues

Ten maternal-fetal pairs were screened to identify informative alleles to quantify the level of maternal microchimerism in DNA extracted from fetal tissues, as described above. Three pairs were excluded from analysis because an informative allele assay could not be identified. The remaining seven fetuses were tested for maternal microchimerism in the thymus, spleen, liver, mesenteric lymph nodes, and bone marrow. In some cases, more than one informative allele was identified. In each case where maternal microchimerism was detected, direct sequencing of the PCR product (for informative allele assays yielding a product longer than 100 bp) or cloning followed by sequencing (for short amplicon assays) was used to confirm the 100% identity of the PCR product detected in the fetal sample relative to the product amplified from the maternal sample. Furthermore, in some cases, paternal DNA was also amplified and sequenced in parallel, using the informative allele assays to rule out the possibility that weak amplification of the inherited paternal allele in the fetal DNA was misinterpreted as maternal microchimerism. In those cases, paternal alleles were clearly distinguishable from maternal alleles by at least six nucleotide mismatches (data not shown).

A summary of maternal microchimerism levels in fetal tissues is presented in Table 1. Two of seven tested had detectable levels of microchimerism in at least one of the tissues analyzed (Table 1): in one (#2), microchimerism was detected in all five tissues with a range of 0.005% to 0.07%; in another (#6), a low-level positive result near the limit of detection was identified in the bone marrow in one of three independent experiments, each using two different assays, suggesting that a single copy or a small number of copies of the target sequence was present in the aliquots tested for that fetus. Within the limits of detection of the assays (ranging from 0.001% to 0.05%), no maternal microchimerism was detected in any tissue in five of the fetuses. Importantly, a similar number of cells was tested for each sample. Several of the informative allele-specific PCR assays demonstrated cross-amplification to a minor extent of MHC alleles in the fetal DNA other than the targeted non-inherited maternal alleles, despite amplification at relatively high annealing temperatures (Table 1). Sequencing of the amplified products allowed discrimination between maternal-specific DNA and weak cross-reactivity to fetal DNA.

Transient maternal microchimerism levels in rhesus monkey PBMCs

Fifteen rhesus monkey maternal-infant/juvenile pairs were screened for informative alleles that could allow quantification of maternal microchimerism in DNA extracted from PBMCs. After exclusion of three pairs where an informative allele assay could not be identified, 12 pairs were tested for maternal microchimerism in PBMCs. Four offspring had both prenatal and postnatal samples available for microchimerism analyses whereas the remaining eight focused on postnatal samples because the fetal samples were used for other study-related assays. In the early third trimester, fetal blood was collected by cardiocentesis (left ventricle) under ultrasound guidance, an approach that minimizes the potential for maternal blood contamination. At this time point, maternal microchimerism was detected in fetal PBMCs from all four fetuses assessed (0.0236 ± 0.0056%, 0.0153 ± 0.0123%, 0.0011 ± 0.0008%, 0.0389 ± 0.0086% for #8–11, respectively). No microchimerism was detected in umbilical cord blood from these four animals at birth or in PBMCs postnatally (range 4 to 51 weeks). Eight other offspring were tested for maternal microchimerism in PBMCs at birth and subsequent postnatal time points. No microchimerism was detected in these eight infants/juveniles at any of the time points assessed (data not shown).

Persistence of maternal microchimerism in rhesus monkey tissues postnatally

Twelve juveniles assessed for microchimerism in PBMCs were also analyzed for maternal microchimerism in tissues (e.g., thymus, spleen, liver, mesenteric lymph nodes, and bone marrow) at ~1–1.5 years of age. Maternal microchimerism ranging from 0.004% to 1.9% was detected in at least one tissue in three out of 12 evaluated (Table 1). Microchimerism levels and tissue distribution varied widely between each of the animals.

Animals

All animal procedures conformed to the requirements of the Animal Welfare Act and protocols were approved prior to implementation by the Institutional Animal Care and Use Committee at the University of California, Davis. Normally cycling, adult female rhesus monkeys (Macaca mulatta) (n = 25) with a history of prior pregnancy were selected based on negative testing for Mamu-A*01, A*02, B*08, and B*17 alleles, and bred with males that were positive for Mamu-A*01 and A*02 alleles and negative for B*08 and B*17 alleles. Pregnancies were established by timed mating and detected by ultrasound.³⁹ Pregnancy in the rhesus monkey is divided into trimesters by 55-day increments, with 0–55 days of gestation representing the first trimester, 56–110 days of gestation representing the second trimester, and 111–165 days of gestation the third trimester (term 165 ± 10 days).⁴⁰ Activities related to animal care (e.g., diet and housing) were performed according to California National Primate Research Center standard operating procedures (SOPs). Fetal growth and development were monitored by ultrasound, fetal blood was collected by cardiocentesis under ultrasound guidance in the early third trimester using established techniques,³⁹ and maternal blood samples were collected during gestation and at pregnancy termination. Hysterotomies were performed (n = 10) to collect fetal tissues, using established protocols⁴¹ based on study assignment, and included controls (n = 4) or fetuses of dams that had been administered nonpathogenic SIVmac1A11 by the intraperitoneal (IP) route in the late first trimester (n = 6). Fetal body weights and measures were assessed after the collection of fetal blood, then all organs including the brain, lung, heart, thymus, spleen, liver, lymph nodes (axillary, inguinal, mesenteric), pancreas, adrenals, kidneys, reproductive tract including gonads, gastrointestinal tract (stomach, duodenum, jejunum, ileum, colon), skin, muscle, and bone marrow were removed and select organs weighed. A placental evaluation was also performed, and sections of the umbilical cord, membranes, decidua, and placenta obtained. Cell suspensions from thymus, liver, spleen, lymph nodes, and bone marrow were prepared according to established protocols.⁴¹

Cesarean sections were also performed at term (n = 15) using established methods and included collection of umbilical cord blood and a placental evaluation.⁴² Newborns were raised in the nursery according to established SOPs, with blood samples collected from a peripheral vessel at defined time points (weekly or monthly) under ketamine sedation (10 mg/kg).⁴³ PBMCs were purified from whole blood (of the dam, fetus, or infant/juvenile) by density centrifugation over a Histopaque gradient (Sigma, Catalog #10771). Isolated cells were subsequently cryopreserved in fetal bovine serum (FBS) supplemented with 10% dimethyl sulfoxide (DMSO) prior to DNA extraction. Of these 15 animals, seven were fetal controls and received either pathogenic SIVmac239 or SIVmac251 postnatally (as part of a parallel series of studies). For five, either the dam (intravenous) or the fetus (IP) was administered nonpathogenic SIVmac1A11 in the late first trimester, and three were administered a lentiviral vector prenatally (IP) and SIVmac251 postnatally. Tissue harvests were performed according to established protocols at ~1 to 1.5 years postnatal age and five collected tissues were analyzed from each animal (thymus, liver, spleen, mesenteric lymph nodes, and bone marrow). Tissues were collected and weighed as described above, and cell suspensions prepared according to established protocols as noted.

DNA extraction for PCR genotyping

An estimated 5x10⁶ cells were washed in 1 ml PBS and digested in 500 μl of PCR solution (100 mM KCl, 10 mM Tris pH 8.3, 2.5 mM MgCl₂, 1% Tween-20, 1% NP40) and 75 μg of protease (Qiagen Inc., Cat#19155) at 60 °C for 1.5 h. Protease was inactivated by heating at 95 °C for 30 min. For typing reactions, genomic DNA was diluted 1:50 in PCR solution to a concentration of approximately 1000 genomic equivalents/5 μl.

Design and development of Mamu-MHC allele-specific real-time PCR assays

Four A locus primers were designed based on polymorphic regions in published Mamu-A gene sequences.⁴⁴ Likewise, 10 B locus primers were designed based on polymorphic regions in Mamu-B gene sequences included within the Immuno Polymorphism Database.^45,46 All A and B locus primers are located within the highly variable second and third exons. The A locus primers were combined to form two assays and the B locus primers were combined into six assays, each of which targets a single MHC class I allele. In addition, we also used techniques kindly provided by Dr. David Watkins, Department of Pathology, University of Miami Miller School of Medicine, as well as published assays,^47-49 to form a panel consisting of 12 MHC allele-specific qPCR assays and one GAPDH reference assay used as a loading control. In addition to the MHC allele-specific assay panel, we performed sequence-based typing of Mamu-DQB for each animal in these studies and designed qPCR assays specific for the non-inherited DQB maternal allele for 12 maternal-fetal pairs. The real-time PCR primer sequences and amplicon characteristics (e.g., size and dissociation temperature) are shown in Table 2.

To validate the utility of the MHC allele-specific assay panel for detection of microchimerism, the assays were used to screen rhesus monkey DNA from at least five different animals. As shown by the melting curve analysis from one representative assay for Mamu-A1*003 (Figure 2A), the primer pairs were informative for distinguishing between rhesus monkey samples known to be positive or negative for this allele. The predicted sizes of the amplified products were verified on an ethidium-stained agarose gel (data not shown). Amplification of serial dilutions of DNA with the Mamu-MHC allele-specific assays showed linearity over at least four orders of magnitude. The linear dynamic range for the reference assay for GAPDH and a representative allele-specific MHC PCR assay (designed to target B*047) are shown in Figure 2B. To determine the limit of detection of maternal DNA in a fetal sample, serial dilutions of rhesus macaque DNA positive for the targeted MHC allele were spiked into a fixed amount of DNA negative for the targeted allele. The relative amount of DNA in each sample was determined by GAPDH qPCR using the 2^-ΔCt method. The sensitivity of the Mamu-MHC allele-specific PCR assays, defined as the lowest amount of DNA positive for the targeted MHC allele detected in a background of DNA negative for the targeted allele, ranged from 0.001% to 0.05% (data not shown).

Identification of informative alleles for maternal microchimerism by real-time PCR

To determine which assays were informative for detection of maternal microchimerism, the MHC allele-specific qPCR panel was used to genotype samples from maternal-fetal rhesus monkey pairs. Five μl of the diluted DNA lysate were added to 10 μl of PCR buffer containing 5 mM MgCl₂, 1 mM dNTPs (Bioline USA Inc., Catalog #39029), 1 μM of each primer (Integrated DNA Technologies), 0.25X SYBR Green I (Invitrogen, Catalog #S7585), and 0.7 U FastStart Taq (Roche Applied Science, Catalog #04738420001). Real-time PCR was performed in a 384-well format on a sequence detection system (LightCycler 480 II, Roche) with the following cycle conditions: 1 min at 95 °C followed by 45 cycles of 30 s at 95 °C, 30 s at 60 °C, and 45 s at 72 °C, with a melting curve analysis at the end of the reaction. Any weak maternal-specific signal detected in the fetus was considered as possibly derived from microchimeric cells that crossed the placenta from the dam into the fetus. An assay was defined as “informative” for maternal microchimerism if a Ct between 20 and 30 was detected in the maternal sample and if the Ct of the fetal sample was either undetectable or at least five cycles higher than that of the maternal sample for equivalent maternal and fetal DNA input. Examples of informative and non-informative alleles identified by qPCR genotyping are presented in Figure 2C, i and ii, respectively. Once informative alleles were identified, 10-fold serial dilutions of the DNA lysate were amplified in a 96-well format using different annealing temperatures ranging from 56–70 °C. Optimal annealing temperatures were chosen for each assay based on increased specificity at higher temperatures without a loss in sensitivity.

PCR analysis and quantification

To quantify the level of maternal microchimerism, fetal DNA and 10-fold serial dilutions of maternal DNA samples diluted in PCR solution were amplified with primers for informative allele(s) and GAPDH. Quantification of maternal microchimeric cells in fetal DNA was calculated using the ΔCt method with GAPDH as a reference gene. The specificity of the amplified products in the fetal samples was determined by melting curve analysis and gel electrophoresis in comparison to amplified products in the maternal samples.

Sequencing of PCR products

Thirty-five μl of the PCR product were electrophoresed in a 2% agarose gel containing ethidium bromide. The band was excised and purified using a QIAquick PCR purification kit (Qiagen, Catalog #28106). The purified PCR product was eluted into 30 μl of water and 20 μl was submitted with the 5′ and 3′ primers used during amplification for automated sequencing (MCLAB). Alternatively, the PCR product was directly purified using a QIAquick PCR purification kit prior to sequencing with the amplification primers. For amplicons shorter than 100 bp, the PCR product was cloned into pCR^TM4-TOPO^® using the TOPO^® TA Cloning^® kit for sequencing (Life Technologies, Catalog #K4575) and for transformation of One Shot^® TOP10 chemically competent E. coli. Transformants were analyzed by PCR using M13 Forward(-20) and M13 Reverse primers and visualized by agarose gel electrophoresis. Amplicons of the correct size were purified using a QIAquick PCR purification kit prior to sequencing with the M13 primers. Sequence-based typing of Mamu-DQB1 was performed by amplifying exon 2 (270 bp) using the primers DQBF (5′-TCCCCGCAGA GGATTTCGTG-3′) and DQB*18R (5′-CGCTCACCTC GCCGCTGCAA-3′) and sequencing the PCR products using the amplification primers. Sequence analysis was performed using Sequencher^® version 5.0 software (Gene Codes Corporation). Mamu-DQB sequences for each maternal-fetal pair were aligned and primers specific for the non-inherited maternal DQB allele were designed such that there was a 1–3 nucleotide mismatch at each primer 3′ end relative to the fetal DQB sequences.