Fetal origins of hematopoietic failure in a murine model of Fanconi anemia

Bone marrow (BM) failure is the most common cause of morbidity and mortality from Fanconi anemia (FA), a recessively inherited disorder resulting from biallelic mutations in 1 of 15 known genes that cooperate in a DNA repair pathway.^1,2  The presumptive postnatal loss of hematopoietic stem and progenitor cells (HSPC) from the marrow is thought to account for patient symptom onset during early school age and is generally consistent with experimental evidence.³  However, cytopenias often precede clinical symptoms in FA patients or affected siblings.⁴  Although systematic studies of in utero development of FA patients have not been conducted, reports of reduced cord blood progenitor frequency, evidence in embryonic stem cell lines, and studies of rare fetal tissues all support the notion of early compromise in HSPC function.^5-10  Here, we studied fetal liver (FL) hematopoiesis in a murine model of FA complementation group C (Fancc^−/−) to determine if hematopoietic deficits were limited to HSPC losses from the BM compartment, or whether they may have origins earlier in development.¹¹  Our study for the first time reveals broad defects of hematopoietic development in Fancc^−/− animals, which include a compromised fetal liver HSPC pool as well as late gestational lethality.

Mice were handled in accordance with the Oregon Health and Science University Institutional Animal Care and Use Committee. FL from 14.5 days post coitum timed pregnancies (vaginal plug method) from C57BL/6 Fancc^+/− (CD45.2) dams were dissected to prepare single-cell suspensions. Fetal tails were genotyped using the Phire PCR Kit (Thermo Scientific, Waltham, MA). BM cells from adult animals were harvested as previously described.¹² 

Cytogenetic analyses were conducted on FL cells by the Oregon Health and Science University Cytogenetics Core Laboratory. Methylcellulose progenitor assays (M3434; StemCell Technologies, Vancouver, BC, Canada) were performed according to manufacturer’s instructions.

Unfractionated FL cells were cultured in StemSpan SFEM (StemCell Technologies) at 1.6 × 10⁷ cells/mL. Conditioned media was applied to Mouse Cytokine Antibody Array, Panel A (R&D Systems, Minneapolis, MN) following manufacturer’s instructions and imaged on a Fujifilm LAS2000. Pixel density was analyzed using Multi Gauge software (Fujifilm, Tokyo, Japan).

Unfractionated FL cells (2 × 10⁶) were injected via tail vein into irradiated (530 cGy) congenic CD45.1 recipients. Serial chimerism was analyzed as described.¹²  For 2° transplant, 2 × 10⁶ whole BM cells from 1° recipients (n = 2/genotype) were transplanted (each) into 5 irradiated (775 cGy) CD45.1 recipients.

For c-Kit⁺ Sca-1⁺ Lin⁻ (KSL) and AA4.1⁺ Sca-1⁺ Lin^low/− (ASL) HSPC analyses, cells were stained using antibodies; lineage antibodies were against B220, Gr-1, CD3, CD4, CD5, and Ter119 (excluding Mac-1¹³ ). For cell-cycle analysis, we used anti-Ki-67 antibody and Hoechst 33342.¹⁴  Apoptosis assays used Annexin V (BD) and reactive oxygen species (ROS) levels were assayed by carboxy-H₂DCFDA staining (Invitrogen, Carlsbad, CA). Data were collected on FACSCalibur and BD Influx (BD Biosciences, San Jose, CA) and analyzed using FlowJo software (TreeStar, Ashland, OR).

KSL or ASL FL cells were sorted by FACS on a BD Influx. RNA was purified using RNeasy Mini Kit (Qiagen, Germantown, MD), and then converted to complementary DNA using SuperScript III First Strand (Invitrogen). The complementary DNA was analyzed by real-time polymerase chain reaction (PCR) (StepOne Plus) using Power SYBR Green Master Mix (both Applied Biosystems, Carlsbad, CA). Fancc^−/− and wild-type (WT) samples were normalized to Gapdh, and Fancc^−/− gene expression was normalized to WT. Relevant transcripts were previously described.^9,15 

Numerical results are expressed as mean (± standard error of the mean as indicated) and compared using an independent Student t test. Genotype distributions were compared by χ-square test.

A wide range of FA-associated constitutional defects arising in utero are diagnosed at birth. Hematopoietic manifestations of FA in infants, by comparison, are rare.⁴ Fancc^−/− mice provide a validated, if imperfect, model of FA hematopoiesis. Here, we focused on FL hematopoiesis (14.5 days post coitum) as a key stage during hematopoietic development.¹⁶  Using heterozygous breeder animals, the genotype distribution across 28 litters yielded 26% Fancc^−/− fetuses (n = 59^−/−, 81^+/+, 87^+/−), at near-Mendelian frequency. This exceeds the frequency at weaning of 18.9% (n = 659 total; P = .004), also reported by others, and for the first time reveals late gestational lethality in FA mice (Figure 1A).^17,18  Beyond occasional gross morphological abnormalities (microphthalmia, anophthalmia, anencephaly), Fancc^−/− fetuses weighed on average 8% less than WT littermates, with a concomitant 34% reduction in whole FL cellularity and a 15% decrease in placental mass (Figure 1B-H). Fancc^−/− FL cells exhibited characteristic mitomycin-c–induced radial formation (Figure 1I) and chromosomal breaks (not shown).^11,19  Clonogenic hematopoietic progenitor growth in methylcellulose was reduced by 20% in Fancc^−/− FL compared with WT (Figure 1J), implying an aggregate 47% decline in progenitor frequency per fetus. Although interleukin-1, tumor necrosis factor-α, and interferon-γ can suppress postnatal hematopoiesis in Fancc^−/− animals, we found no differences in cytokine secretion by unfractionated FL cells to account for the reduction in fetal Fancc^−/− progenitor cells (Figure 1K).²⁰  Our data reveal other phenotypic differences between fetal FA HSPC and those present in the postnatal bone marrow compartment. Unlike adult FA HSPC, characterized by exaggerated apoptosis and elevated levels of ROS, we found no difference in the frequency of apoptotic events (Annexin V and expression of pro-apoptotic genes Puma and Noxa) or ROS levels (by carboxy-H₂DCDFA stain) between Fancc^−/− and WT Sca-1⁺ AA4.1⁺ (fetal HSPC marker²¹ ) FL cells (Figure 2A-C).^22,23  Cell-cycle analysis revealed fewer Fancc^−/− than WT c-Kit⁺ Sca-1⁺ (KS) Ki-67^neg (G₀) HSPC (Figure 2D), echoing the reported loss in quiescence in murine FA BM KSL cells, and suggested potential G₁ arrest and senescence.^14,19,24  This is further consistent with quantitative reverse-transcription-PCR studies of sorted FL HSPCs demonstrating Fancc^−/− upregulation of p21 and p18, cell-cycle regulators involved in hematopoietic failure in FA (Figure 2E).^6,9 

We previously showed that the frequency of immunophenotypically defined HSPC (c-Kit⁺ Sca-1⁺ Lin⁻; KSL) in Fancc^−/− BM is not substantially different from WT animals.¹²  Similarly, we found a minimal decrease in Fancc^−/− FL AA4.1⁺ Sca-1⁺ Lin^low/− (ASL) HSPCs (Figure 2G). Deficient repopulation is widely considered a hallmark of compromised FA HSPC function.^8,24  We therefore tested serial repopulating ability by transplanting 2 × 10⁶ CD45.2 Fancc^−/− or WT unfractionated FL cells into sublethally irradiated, congenic CD45.1 mice. Peripheral blood donor chimerism 5 months after transplantation showed an average genotype difference of 24% (P = .14) (Figure 2F-H). However, during serial repopulation, the average donor chimerism between 1° and 2° recipients decreased from 87% to 66% in WT, compared with 66% to 20% for Fancc^−/− donor cells (21% vs 46% decline, respectively) (Figure 2I). Thus, the average difference in chimerism between genotypes grew from 24% in 1° recipients to 70% in 2° recipients (P = .0002 for 2° transplant).

Fancc^−/− mice do not demonstrate the slowly progressive cytopenias seen in patients, but they have provided valuable insight into FA HSPC biology. Our data reveal that the hematopoietic defect in Fancc^−/− mice has origins in development without compartmental or temporal restriction to the BM. These observations provide the cohesive developmental context for prior reports that suggested prenatal HSPC loss and have potential implications for ongoing efforts to harness induced pluripotency for FA.^5-7  It will be important to delineate how the FA phenotype impacts HSPC pool expansion and seeding of the fetal BM niche during development. Finally, compromised prenatal hematopoiesis may determine the timing of diagnostic procedures, a reassessment of its mechanistic origin, and prompt strategies to preempt hematopoietic failure in FA patients.

The authors thank Dr. Amy Skinner, Dr. Jianya Huan, Tae Hoon Ha, Kyle Lenz, Andrea McBeth, Pamela Canaday, Dorian La Tocha, and the Oregon Health and Science University Cytogenetics Core.

A.N.K.-L. was supported by T32 Molecular Hematology Training Grant HL778116 and N.L. Tartar Trust Grant. Presented in part at the 23rd Annual Fanconi Anemia Research Fund Scientific Symposium, Barcelona, Spain, 2011, and the 53rd Annual Meeting of the American Society of Hematology, San Diego, CA, 2011.

Contribution: A.N.K.-L. designed and performed experiments and wrote the manuscript; N.A.G. performed experiments and interpreted data; P.K. designed experiments, interpreted data, and wrote the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Peter Kurre, Oregon Health & Science University, Papé Family Pediatric Research Institute, L499, 701 SW Gaines Rd, Portland, OR 97239-3098; e-mail: kurrepe@ohsu.edu.