Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jun 13.
Published in final edited form as: Blood. 2026 Feb 26;147(9):973–986. doi: 10.1182/blood.2025028895

p27Kip1 regulates γ-globin production

Ginette Balbin-Cuesta 1,2, Claire Drysdale 2,3, Claire Kerpet 4, Lei Yu 4, Greggory Myers 3, Zesen Lin 5, Beth McGee 4, Ann Friedman 4, Xiaofang Liu 6, Sharon A Singh 6,7, James Douglas Engel 3,7, Laura Buttitta 8, Rami Khoriaty 1,3,4,7,*
PMCID: PMC12957197  NIHMSID: NIHMS2179729  PMID: 41411145

Abstract

Sickle cell disease (SCD) and β-thalassemia are devastating genetic disorders resulting from defects in the β-globin subunit of adult hemoglobin (HbA). Both disorders are ameliorated by the induction of γ-globin, a component of fetal hemoglobin (HbF). Therefore, the development of safe, effective, and widely available inducers of HbF is needed. Here, we discovered that slow cycling erythroid cells exhibit increased γ-globin expression. To understand the molecular basis of this, we screened all cyclin-dependent kinase inhibitors (CDKIs) for their ability to induce HbF using CRISPR activation (CRISPRa). We found that overexpression of CDKN1B, which encodes p27Kip1 (but not overexpression of other CDKIs), induces γ-globin expression at the transcriptional level. CDKN1B mutants expressing proteins unable to bind/inhibit CDKs and/or cyclins revealed that γ-globin induction by p27Kip1 depends largely on the domains involved in its cell cycle function. Pharmacological inhibition and genetic reduction of CDK4/6 also results in increased HbF. In genetic rescue experiments, we show that p27Kip1 induces HbF by inhibiting CDK4/6, through a mechanism that is likely BCL11A and ZBTB7A independent. Furthermore, palbociclib, an oral CDK4/6 inhibitor, significantly increases HbF in a murine SCD model at doses that are well tolerated. Moreover, we show that HbF induction by hydroxyurea, a drug currently in use to treat SCD, may be mediated in part by CDK4/6 inhibition. Overall, our findings establish a causal relationship between CDK4/6 activity and γ-globin production and suggest that single or dual CDK4/6 inhibitors might be therapeutically beneficial for SCD and β-thalassemia.

Introduction

Sickle cell disease (SCD) and β-thalassemia affect millions of individuals worldwide and pose a mounting public health burden1,2. These inherited disorders result from qualitative or quantitative defects in the β-globin subunit of adult hemoglobin (HbA)1,2. Upregulation of γ-globin, the β-globin subunit of fetal hemoglobin (HbF) that is predominantly expressed during fetal development, has been shown to ameliorate both SCD and β-thalassemia1,2. This is due, at least in part, to γ-globin’s ability to inhibit hemoglobin polymerization in SCD and functionally compensate for reduced/depleted β-globin subunits in β-thalassemia3. Based on these observations, the development of accessible treatment modalities that upregulate γ-globin production is needed.

Cell cycle dynamics play a major role in the proper progression of erythroid differentiation4. Early in differentiation, the fate of megakaryocyte-erythroid progenitors (MEPs) is influenced by cell cycle speed. Faster cycling MEPs are biased toward an erythroid fate, while slower cycling MEPs are biased toward a megakaryocytic fate5. Following erythroid commitment, the earliest fully committed erythroid cells, known as burst forming unit-erythroid (BFU-Es) cells, exhibit slow cell cycles with a high G1 to S phase ratio4,6. BFU-Es generate colony forming unit-erythroid (CFU-Es) cells via a process associated with shortening of G1, resulting in a higher percentage of CFU-Es in the S phase of the cell cycle7,8. The transition from CFU-E to terminal erythroid differentiation occurs during a notably short S-phase due to downregulation of the cyclin-dependent kinase inhibitor (CDKI) p57Kip2 7,8. Subsequently, during terminal erythroid maturation, the cell cycle speed slows gradually4, 9. These findings demonstrate an intricate interconnection between the cell cycle and erythroid differentiation. Additional studies have shown that other cell cycle regulators play critical roles in controlling erythroid cell size and number10,11, coordinating erythroid cell cycle exit and enucleation12–17, and modulating erythroid differentiation18,19.

Intriguingly, several small molecules that have been shown to affect γ-globin expression also affect the erythroid cell cycle20,21. One notable example of such molecules is hydroxyurea, which increases γ-globin expression and improves morbidity and mortality of patients with SCD22,23. However, the precise mechanism of action by which hydroxyurea induces γ-globin expression remains unknown22. Additionally, the relationship between cell cycle regulators and γ-globin production has not been definitively established.

In this report, we show that slow cycling erythroid cells exhibit increased γ-globin expression. We reveal that overexpression of the CDKI p27Kip1 (but not other CDKIs) leads to induction of γ-globin expression, with the latter mediated by CDK4/6 inhibition. Additionally, we show that hydroxyurea appears to increase F-cell production by CDK4/6 downregulation, which may explain, at least in part, the mechanism of its therapeutic value in SCD. Our findings illustrate a causal link between cyclin dependent kinase activity and γ-globin production and support the additional consideration of CDK4/6 inhibitors (such as palbociclib) as possible therapeutic strategies for treating β-hemoglobinopathies.

Methods

CFSE Labeling

CFSE (Thermo Fisher, catalog # C34554) was incubated with cells at various stages of erythroid differentiation, per manufacturer’s instructions. Forty-eight hours post-CFSE labeling, cells were stained for intracellular HbF levels as described in the supplement.

RT-qPCR and Western blot

RNA isolation and qRT-PCR were generally performed as previously described24,25, as was cell lysate collection and western blotting26,27. However, these methods are described in more details in the supplement. Primers used for qRT-PCR are listed in Table S2.

Townes SCD Mice

The previously described28,29 Townes SCD humanized mouse model was obtained from the Jackson Laboratory. These mice were treated with Palbociclib or vehicle control as described in the supplement.

Complete blood count analysis

Blood was drawn from mice via submandibular venipuncture and then collected into RAM Scientific Safe-T-Fill Capillary Blood Collection Systems tubes coated with EDTA (Fisher cat. #14-915-50). Complete blood count analysis was done using the Advia 120 Hematology System (Siemens), as previously described30,31.

Other methods

Detailed methods for HUDEP-2 cell culture, differentiation of erythroid cells from HSPCs, cloning and lentiviral preparation, transduction of HUDEP-2 and CD34+ cells, cell staining and sorting, RT-qPCR, western blotting, antibodies used for flow cytometry and western blotting, drug treatment, high-performance liquid chromatography, and cytospin preparation are included in the supplement.

Results

Correlation between the cell cycle speed and % F-cell formation

To evaluate whether cell cycle speed correlates with γ-globin expression, we labeled HUDEP-2 cells (which primarily express β-globin32) with carboxyfluorescein diacetate succinimidyl ester (CFSE) at different days of differentiation. With each cell division, CFSE intensity is expected to be reduced by half; therefore, faster cycling cells retain less CFSE than slower cycling cells. Starting at Day 4 of differentiation, the slowest cycling (top 10% CFSE) HUDEP-2 cells exhibited an increased proportion of γ-globin expressing cells (F-cells) compared to the fastest cycling (bottom 10% CFSE) cells, reaching a ~four-fold difference at day 8 of differentiation (Fig 1A-C).

Figure 1. Cells cycling at slower speeds exhibit increased F-cell formation.

Figure 1.

Open in a new tab

(A) A schematic demonstrating that HUDEP-2 cells stained with CFSE were analyzed 48 hours later for CSFE fluorescence and γ-globin expression by flow cytometry at different days of differentiation. (B) Representative flow plots demonstrating percent F-cells at day 8 of differentiation in the slowest (top 10% CFSE) and fastest (bottom 10% CFSE) cycling cells. (C) Percent F-cells in slowest and fastest cycling HUDEP2 cells at various days of differentiation (n=3). (D) A schematic of the differentiation strategy of CD34+ HSPCs into erythroid cells in vitro. CFSE staining and FACS analysis for CSFE fluorescence and γ-globin expression were performed at differentiation days 13 and 15, respectively. (E) Percent F-cells among HSPC-derived erythroid cells at day 15 of differentiation. Each experiment was done in technical triplicates, using HSPCs derived from 3 independent donors. p-values in panel C were calculated via 2-way ANOVA, while p-values in panel E were calculated using unpaired t-test. All data are shown as mean +/− SD.* p<0.05; ** p<0.01; *** p<0.001; **** p<0.0001.

To exclude the possibility that the above observations might be an artifact of the HUDEP-2 cells, we differentiated primary human CD34+ hematopoietic stem and progenitor cells (HSPCs) into erythroid cells in vitro, incubated cells with CFSE at day 13 of differentiation, and analyzed the proportion of F-cells at day 15 of differentiation (Fig 1D). Using data from three independent HSPC donors, the slowest cycling primary erythroid cells exhibited a significant increase in F-cell formation compared to the fastest cycling erythroid cells (Fig 1E). Taken together, these data illustrate a correlation between slow cell cycle speed and increased F-cell formation.

Impact of CRISPRa-mediated CDKI overexpression on γ-globin expression

To determine if there is a causal relationship between cell cycle speed and F-cell formation, we overexpressed each of eight CDKIs by CRISPR activation (CRISPRa) and evaluated the impact of each CDKI overexpression on γ-globin expression.

We first generated a clonal HUDEP-2 cell line that stably expresses the transcriptional activator complex MPH (MS2-P65-HSF1). We validated the capacity of this cell line, referred to as HUDEP-2-MPH, to adequately differentiate and up-regulate gene expression using two independent sgRNAs targeting HBG (Fig S1A-B). Subsequently, we transduced the HUDEP-2-MPH cell line with lentiviral particles (lentiSAMv2) that express sgRNAs targeting one of eight CDKIs (CDKN1A, CDKN1B, CDKN1C, CDKN2A, CDKN2B, CDKN2C, CDKN2D, and CDKN3) (Fig 2A). Transduced cells were selected with blasticidin and subsequently differentiated. mRNA was collected at day 3 of differentiation to validate overexpression of target genes, and cells were analyzed for F-cell formation on day 8 of differentiation (Fig 2B).

Figure 2. CDKI overexpression in HUDEP-2 cells uncovers a role for p27Kip1 in γ-globin expression.

Figure 2.

Open in a new tab

(A) An illustration showing the 8 CDKIs targeted by CRISPRa. (B) Experimental timeline depicting transduction of HUDEP-2-MPH cells with CRISPRa sgRNAs, followed by antibiotic selection for 5 days, HUDEP-2 cell differentiation, and analysis at differentiation days 3 and 8. (C) Fold change in counts of differentiated HUDEP-2-MPH cells following transduction with sgRNAs targeting one of eight CDKIs or non-targeting (NT) control sgRNAs. (D) F-cell percentage among HUDEP-2 cells overexpressing one of 8 CDKIs (using CRISPRa) at day 8 of differentiation. (E) RT-qPCR for CDKN1B in HUDEP-2-MPH cells transduced with CDKN1B-targeting or NT sgRNAs. (F) RT-qPCR for HBG1/2 in HUDEP-2-MPH cells transduced with CDKN1B-targeting or NT sgRNAs. p-values in panel C are shown for cells at differentiation day 8 and are calculated via 1-way ANOVA. p-values for panel D were calculated via 1-way ANOVA. p-values for panels E and F were calculated using unpaired t-test. All data are shown as mean +/− SD. ** p<0.01; *** p<0.001; **** p<0.0001.

Transduction of HUDEP-2-MPH cells with lentiSAMv2 resulted in increased expression of target genes as measured by RT-qPCR (Fig S1C). All CDKI-targeting sgRNAs resulted in a reduction in cell growth over time compared to cells transduced with non-targeting (NT) sgRNAs, an expected finding due to reduced cell cycle speed (Fig 2C and Fig S1C).

Intriguingly, the only CDKI whose activation resulted in a statistically significant increase in F-cell formation was CDKN1B, which encodes p27Kip1. Compared to cells transduced with NT sgRNAs, HUDEP-2-MPH cells transduced with CDKN1B-targeting sgRNAs exhibited a ~6-fold increase in the proportion of F-cells (Fig 2D). We next found an ~11-fold increase in γ-globin mRNA levels in cells transduced with CDKN1B-targeting versus NT sgRNAs (Fig 2E-F).

Taken together, these findings suggest that overexpression of CDKN1B results in increased γ-globin expression at the transcriptional level. However, slowing cell cycle speed is not sufficient by itself to result in increased F-cell formation, as overexpression of CDKIs other than CDKN1B failed to induce γ-globin.

p27Kip1 overexpression results in increased γ-globin production in HUDEP-2 cells

To exclude the possibility of an off-target effect resulting from the CDKN1B targeting sgRNAs, we overexpressed CDKN1B using a lentiviral expression construct that also expresses a blue fluorescent protein (CDKN1B-mTagBFP2) (Fig 3A). Compared to control cells expressing mTagBFP2 only (Empty Vector (EV)), HUDEP-2 cells transduced with CDKN1B-mTagBFP2 lentivirus exhibited a gradual reduction in the proportion of transduced cells (assessed by % mTagBFP2 positive cells by flow cytometry) over time (Fig 3B). This finding is a consequence of reduced cell cycle speed mediated by CDKN1B overexpression, leading to outcompetition by cells that were not transduced with the CDKN1B-mTagBFP2 lentivirus.

Figure 3. p27Kip1 overexpression in HUDEP-2 cells results in increased γ-globin production.

Figure 3.

Open in a new tab

(A) Vectors overexpressing CDKN1B and mTagBFP2 (CDKN1B-mTagBFP2) or mTagBFP2 only (empty vector (EV)) were generated and transduced into HUDEP-2 cells by spinfection followed by data analysis at day 8 of differentiation. (B) Percent mTagBFP2 positive cells was measured by flow cytometry every 2 days from days 2–8 of differentiation. (C) RT-qPCR for CDKN1B and (D) percent F-cells in p27Kip1 overexpressing and control HUDEP-2 cells at day 8 of differentiation. (E) RT-qPCR for HBG1/2 and HBB at day 8 of differentiation. (F) Representative flow cytometry plots demonstrating erythroid differentiation analyzed by BAND3 and CD49d expression. Technical triplicates are shown in the bar graph. (G) Representative flow cytometry plot demonstrating percent F-cells in stage matched and size matched erythroid cells, with (H) technical triplicates summarized in a bar graph. p-values in panel B were calculated using 2-way ANOVA and p-values in panels C-F & H were calculated using unpaired t-tests. All data are shown as mean +/− SD. ** p<0.01; *** p<0.001; **** p<0.0001.

HUDEP-2 cells transduced with CDKN1B-mTagBFP2 lentivirus demonstrated increased CDKN1B mRNA levels by RT-qPCR (Fig 3C). Notably, compared to control cells, CDKN1B-overexpressing HUDEP-2 cells exhibited a ~4-fold increase in the proportion of F-cells (Fig 3D) and a ~12-fold increase in γ-globin mRNA (with reduced β-globin mRNA) levels by RT-qPCR (Fig 3E) at day 8 of differentiation.

We next analyzed the impact of CDKN1B overexpression on HUDEP-2 differentiation, as we previously described33. Erythroid cells upregulate BAND3 and downregulate CD49d expression during differentiation25,34. CDKN1B overexpression did not delay this process; instead, cells with increased CDKN1B expression appeared to exhibit a mild increase in BAND3 expression (Fig 3F).

Since CDKN1B overexpression results in larger cells, we gated on mTagBFP+, differentiation-matched, and size-matched cells. We found that increased CDKN1B expression leads to increased F-cell formation, even when controlling for cell size (Fig 3G-H). Thus, increased F-cell formation resulting from CDKN1B overexpression is not due to larger cell size.

p27Kip1 overexpression results in increased γ-globin production in primary human erythroblasts

To exclude the possibility that CDKN1B overexpression and increased F-cell formation is an artifact of the immortalized cell line HUDEP-2, we differentiated HSPCs into erythroid cells in vitro, using a three-phase differentiation protocol (Fig 4A). At day 7 of differentiation, cells were transduced with CDKN1B-mTagBFP2 lentivirus or control mTagBFP2 lentivirus.

Figure 4. p27Kip1 overexpression results in increased γ-globin production in primary human erythroblasts.

Figure 4.

Open in a new tab

(A) Schematic of the 3-phase differentiation strategy of CD34+ HSPCs into erythroid cells in vitro. Spinfection, EdU analysis, and FACS/ RT-qPCR analysis were performed at differentiation days 7, 11, and 15, respectively. (B) Percent mTagBFP2 positive cells measured by flow cytometry in differentiating HSPCs transduced with CDKN1B-mTagBFP2 or control mTagBFP2 lentiviral vectors. Analysis was performed at differentiation days 9, 11, and 15. (C-D) Distribution of cell cycle stages in HSPC-derived erythroid cells at day 11 of differentiation. This experiment was done in triplicates in HSPCs harvested from 2 independent donors. (E) CDKN1B, HBG1/2, and HBB expression by RT-qPCR in HSPC-derived erythroid cells at day 15 of differentiation. This experiment was done in triplicates in HSPCs harvested from one donor and (F) repeated in erythroid cells derived from an independent HSPC donor. (G) % F-cells by flow cytometry among erythroid cells at day 15 of differentiation derived from one HSPC donor and (H) from a second, independent, HSPC donor. (I) Representative flow cytometry plots demonstrating erythroid differentiation (at day 15) analyzed by BAND3 and CD49d expression. Experiments using 2 independent HSPC donors (n=3 per HSPC donor) are shown in the bar graph. p-values in panel B were calculated using 2-way ANOVA. p-values in panels D-I were calculated using unpaired t-test. Data are shown as mean +/− SD. * p<0.05; ** p<0.01; *** p<0.001; **** p<0.0001.

As expected, cells transduced with CDKN1B-mTagBFP2 lentivirus were outcompeted by non-transduced cells over time (Fig 4B). Additionally, a higher proportion of erythroid cells overexpressing CDKN1B were in the G0/G1 phase of the cell cycle compared to control cells (~70% vs. ~48%, respectively) (Fig 4C-D).

Erythroid cells differentiated from HSPCs transduced with CDKN1B-mTagBFP2 lentivirus were FACS sorted for mTagBFP2 expression. mTagBFP2+ cells had a ~five-fold increase in CDKN1B mRNA levels by RT-qPCR compared to controls (Fig 4E-F). Notably, in 2 independent experiments, primary erythroid cells overexpressing CDKN1B exhibited a ~two-fold increase in the proportion of F-cells (Fig 4G-H), and an ~eight-fold increase in γ-globin mRNA levels at day 15 of differentiation (with a concomitant decrease in HBB mRNA) (Fig 4E-F).

We next examined the impact of CDKN1B overexpression on erythroid differentiation. mTagBFP2+ erythroid cells overexpressing CDKN1B exhibited slightly increased BAND3 expression and CD49d downregulation, suggesting enhanced differentiation (Fig 4I). Consistent with these observations, CDKN1B overexpressing cells demonstrated adequate erythroid differentiation morphologically, with larger cell size compared to controls, an expected finding due to reduced cell cycle speed (Fig S2A).

While CDKN1B overexpression resulted in increased F-cell formation, we found no notable effect of CDKN1B downregulation (using two independent shRNA constructs) on F-cell formation or erythroid differentiation (Fig S2B-D).

Next, using data from previously published studies, we examined CDKN1B expression in adult erythroid cells. Compared to CFU-Es and proerythroblasts, the proportion of CDKN1B transcripts sharply increases in basophilic and polychromatic erythroblasts and peaks in orthochromatic erythroblasts and reticulocytes35 (Fig S3A). Of all the CDKIs, only p27Kip1 and p19INK4d protein levels were detected in erythroid cells differentiated from HSPCs36, with much higher levels of p27Kip1 compared to p19INK4d, consistent with a role for p27Kip1 in erythropoiesis (Fig S3B).

γ-globin production mediated by p27Kip1 depends on domains involved in its cell cycle function

p27Kip1 has been shown to have cell cycle-independent functions37–39. Thus, we overexpressed three mutant forms of CDKN1B that are unable to bind CDKs (K), cyclins (C), or both (CK) as previously shown37,40 (Fig 5A), and evaluated their relative impact on γ-globin production compared to cells overexpressing wild-type CDKN1B (WT).

Figure 5. γ-globin production mediated by p27Kip1 depends on domains involved in its cell cycle-functions.

Figure 5.

Open in a new tab

(A) Schematic of p27Kip1 showing the canonical cyclin and cyclin-dependent kinase (CDK) binding sites, as well as the sites mutated to create the p27C, p27CK, and p27K mutant constructs, in which Arginine 30 and leucine 32, phenylalanines 62 and 64, and all 4 latter amino acids were replaced with alanine, respectively. (B) Percent mTagBFP2 positive cells among HUDEP-2 cells transduced with WT p27Kip1, mutant p27Kip1, or EV control. (C) Distribution of cell cycle stages in HUDEP-2 cells expressing WT p27Kip1, mutant p27Kip1, or EV control at day 4 of differentiation (n=3/condition). (D) CDKN1B expression by RT-qPCR in sorted mTagBFP2+ HUDEP-2 cells that express WT p27Kip1, mutant p27Kip1, or EV control (n=3/condition). (E) Percent F-cells among mTagBFP2+ HUDEP-2 cells that express WT p27Kip1, mutant p27Kip1, or EV control at day 8 of differentiation (n=3/condition). (F) HBG1/2 expression by RT-qPCR in sorted mTagBFP2+ HUDEP-2 cells at day 8 of differentiation (n=3/condition). (G) Percent mTagBFP2 positive cells among primary erythroid cells expressing WT p27Kip1, mutant p27Kip1, or EV control. (H) Distribution of cell cycle stages in HSPC-derived erythroid cells expressing WT p27Kip1, mutant p27Kip1, or EV control at day 11 of differentiation. (I) RT-qPCR for CDKN1B in sorted mTagBFP2+ primary erythroid cells expressing WT or mutant p27Kip1. (J) Percent F-cells by flow cytometry among day 15 erythroid cells derived from HSPCs isolated from 2 donors. (K) RT-qPCR for HBG1/2 in sorted primary mTagBFP2+ erythroid cells expressing WT or mutant p27Kip1. Experiment in panels G-K were done in triplicates in HSPCs harvested from 2 independent donors. p-values in panels B, G, and H were calculated by 2-way ANOVA. p-values in panel D-F and I-K were calculated using 1-way ANOVA. Data are shown as mean +/− SD.* p<0.05; ** p<0.01; *** p<0.001; **** p<0.0001.

As expected, HUDEP-2 cells overexpressing CDKN1BCK or CDKN1BC were not outcompeted by non-transduced cells, as demonstrated by a relatively constant fraction of mTagBFP2 positive cells over time (Fig 5B). HUDEP-2 cells overexpressing CDKN1BK were moderately outcompeted by untransduced cells, but not to the same extent as cells overexpressing CDKN1BWT, which were significantly outcompeted by untransduced cells (Fig 5B). Overexpression of WT but not mutant forms of CDKN1B resulted in an increased proportion of cells at the G0/G1 phase of the cell cycle (Fig 5C and Fig S4A).

Despite increased WT and mutant CDKN1B expression (Fig 5D), HUDEP-2 cells overexpressing CDKN1BCK, CDKN1BC, or CDKN1BK did not exhibit an increased proportion of F-cells by flow cytometry (Fig 5E), or increased γ-globin mRNA and protein levels by RT-qPCR and western blot, respectively (Fig 5F and Fig S4B-C).

We next examined the impact of overexpressing mutant forms of CDKN1B on γ-globin production in primary human erythroid cells. Expression of WT but not mutant forms of CDKN1B resulted in outcompetition of transduced cells by untransduced cells over time (Fig 5G), and in an increased proportion of cells in the G0/G1 phase of the cell cycle (Fig 5H).

Despite a comparable increase in CDKN1B expression achieved by WT and mutant constructs (Fig 5I), only primary erythroid cells overexpressing CDKN1BWT exhibited a significant increase in the percent of F-cells at day 15 of differentiation (Fig 5J). Overexpressing CDKN1BC and CDKN1BK did not result in increased γ-globin mRNA levels (Fig 5K), while overexpression of CDKN1BCK resulted in mildly increased γ-globin mRNA levels compared to cells expressing mTagBFP2 only. However, the effect of CDKN1BCK overexpression on the γ-globin mRNA level was much weaker than overexpressing WT CDKN1B (Fig 5K).

Taken together, these findings demonstrate that p27Kip1 results in increased γ-globin production primarily via itsinteraction with proteins critical to cell cycle regulation.

Impact of pharmacologic CDK inhibition on γ-globin expression in vitro and in vivo

Since p27Kip1 slows down the cell cycle speed by inhibiting the activity of CDK1, CDK2, CDK4, and CDK637, we pharmacologically inhibited each of these CDK’s in erythroid cells derived from HSPCs (Fig 6A-B). As expected, all CDK inhibitors, and most notably the pan-CDK inhibitor (Flavopiridol), resulted in impaired cell growth in a dose-dependent manner (Fig 6C and Fig S5A-B). In contrast, and as seen with CDKN1B overexpression (Fig 4I), single or combinatorial inhibition of most CDKs resulted in increased BAND3 expression and CD49d downregulation in HSPC-derived erythroid cells at 15 of differentiation (Fig S5C-D), which may indicate enhanced erythroid differentiation. CDK4, CDK6, and CDK4/6 inhibitors resulted in increased percentage of cells in G0/G1 with a concomitant decrease in the proportion of cells in S-phase (Fig 6D), associated with increased EdU median fluorescent intensity (Fig S5E), which corresponds to shortened S-phase length8. In contrast, inhibition of CDK1 and CDK2 led to more marginal changes, with a decreased proportion of cells in G0/G1 and an increased proportion of cells in S-phase at certain doses (Fig S6A).

Figure 6. p27Kip1 results in increased γ-globin production by inhibiting CDK4/6.

Figure 6.

Open in a new tab

(A) Schematic of the 3-phase differentiation strategy of CD34+ HSPCs into erythroid cells in vitro. Drug treatment, EdU analysis, and FACS/ RT-qPCR analysis were performed at differentiation days 7, 11, and 15, respectively. (B) CDK inhibitors used and their targets. (C) Fold change in counts of differentiated erythroid cells treated with various doses of CA224 (CDK4 inhibitor), BSJ-03-123 (CDK6 inhibitor), and palbociclib (CDK4/6 inhibitor). This experiment was done in triplicate using HSPCs harvested from one donor. (D) Distribution of cell cycle stages in HSPC-derived erythroid cells treated with CA224, BSJ-03-123, and palbociclib at day 11 of differentiation. This experiment was done in triplicate using HSPCs harvested from 2 independent donors. (E) Percent F-cells in HSPC-derived erythroid cells treated with hydroxyurea (HU), CA224, BSJ-03-123, palbociclib, and flavopiridol (pan-CDK inhibitor) at day 15 of differentiation. (F-G) Representative HPLC of HSPC-derived erythroid cells treated with different doses of (F) BSJ-03-123 or (G) Palbociclib at day 15 of differentiation, using HSPCs derived from one donor. HbF/HbF+HbA percentages are shown in each panel. (H) Relative percent HbF assessed by HPLC in red blood cells of Townes mice at baseline (day 0, set as 1%) and 21 days following treatment with palbociclib (for 5 days/week) or vehicle control (n=3/condition). (I) Percent F-cells in HSPC-derived erythroid cells following transduction at day 5 with one of two independent shRNAs targeting CDK4, CDK6, or scramble control shRNA. (J) Percent F-cells in HSPC-derived erythroid cells at day 15 of differentiation following transduction with CDK4/6-Puro overexpressing (or empty vector-puro control) lentivirus and CDKN1B-TagBFP2 (or empty vector control) lentivirus at days 5 and 7 differentiation, respectively. (K) Percent F-cells in HSPC-derived erythroid cells at day 15 of differentiation following transduction with EV-Puro, BCL11A-Puro, or ZBTB7A-Puro lentiviral particles at day 5 of differentiation and CDKN1B-TagBFP2 or TagBFP2 lentivirus at day 7 of differentiation. (L) Percent F-cells in HSPC-derived erythroid cells at day 15 of differentiation following transduction with CDK4/6-Puro (or empty vector-Puro) lentivirus at day 5 of differentiation and treatment with hydroxyurea (50μM) or vehicle control starting at differentiation day 7. Experiment in panels I-L were done in HSPCs harvested from 2 independent donors. p-values in panels C and D were calculated by 2-way ANOVA. p-values in panel E and H-L were calculated by 1-way ANOVA. Data are shown as mean +/− SD. * p<0.05; ** p<0.01; *** p<0.001; **** p<0.0001.

Next, we evaluated the impact of CDK inhibition on F-cell formation. CDK4 and CDK6 inhibitors resulted in significantly increased F-cells at most doses used, with similar findings observed with the dual CDK4/6 inhibitor Palbociclib and the Pan-CDK inhibitor Flavopiridol (Fig 6E and Fig S6B-C). Validation studies using HPLC demonstrated a ~3-fold increase in HbF with the CDK6 inhibitor, a ~2-fold increase in HbF with palbociclib (Fig 6F-G and Fig S7A-B), and no notable increase in HbF in cells treated with the CDK4 inhibitor (Fig S7E). Additionally, except for the highest dose of the CDK2 inhibitor used, CDK1 and CDK2 inhibitors resulted in minimal effects on F-cell formation (Fig S6D-E) and no significant HbF changes via HPLC (Fig S7C-E).

To test whether a CDK4/6 inhibitor can induce HbF in vivo, we obtained Townes SCD mice, which express human α-globin, γ-globin, and β-globin in lieu of their murine counterparts, with both β-globin alleles carrying the SCD mutation (HbSS)28,29. Townes SCD mice were treated 5 days/week with 100 mg/kg of palbociclib or vehicle control for 21 days. Palbociclib treatment resulted in a ~2-fold increase in HbF as measured by HPLC (Fig 6H and Fig S8A), with no changes in white blood cell counts, platelet counts, hemoglobin concentrations, and hematocrit levels (Fig S8B). Notably, palbociclib resulted in a similar increase in HbF in Townes mice heterozygous for the SCD mutation (Fig S8C).

Since stress erythropoiesis has been reported to result in increased γ-globin production41–43, we treated Townes SCD (HbSS) mice receiving palbociclib with phenylhydrazine (to induce hemolytic anemia and erythroid stress) or with vehicle control. Phenylhydrazine treatment did not result in a further increase in HbF in mice receiving palbociclib (Fig S8D). Similarly, phenylhydrazine did not increasea HbF in mice heterozygous for HbS (Fig S8E).

p27Kip1 results in increased γ-globin expression by inhibiting CDK4/6

Since pharmacologic inhibition of CDK4 and CDK6 resulted in the most profound γ-globin induction, we next genetically downregulated the expression of CDK4 and CDK6 using 3 independent shRNAs targeting each gene. Downregulation of CDK4 and CDK6 was confirmed by RT-qPCR (Fig S9A) and resulted in a statistically significant increase in F-cell percentage by flow cytometry (Fig 6I) and a ~2-fold increase in HbF by HPLC (Fig S9B-C).

We next determined whether p27Kip1 overexpression results in increased γ-globin production by inhibiting CDK4/6. Therefore, we overexpressed CDKN1B-mTagBFP2 (or mTagBFP2) in HSPC-derived erythroid cells and transduced these cells with lentiviral particles overexpressing CDK4/6 (which we validated by RT-qPCR (Fig S9D)) or with control particles (EV). Notably, the increased F-cell percentage resulting from CDKN1BWT overexpression was completely rescued by CDK4/6 overexpression (Fig 6J). These findings suggest that p27Kip1 overexpression results in increased γ-globin production either via CDK4/6 inhibition, or that CDK4/6 competes with another factor for interaction with p27Kip1 to induce γ-globin production.

CDK6 modulates γ-globin expression via its kinase function

We deleted the endogenous CDK6 gene in HSPC-derived primary human erythroid cells by electroporating a ribonucleoprotein complex consisting of Cas9 protein and a CDK6-targeting sgRNA. CDK6 deletion, confirmed by ICE analysis (Fig S10A), resulted in increased F-cells (Fig S10B) and HBG1/2 mRNA levels (Fig S10C). We next tested if the γ-globin induction mediated by CDK6 deletion is rescued by expression of a kinase mutant CDK6 cDNA (CDK6 K43M) that renders CDK6 catalytically inactive44. The increased % F-cells and HBG1/2 mRNA levels mediated by CDK6 deletion were rescued by cDNA expression of wildtype CDK6 but not CDK6 K43M (Fig S10 B-C), despite equal expression of WT and K43M mutant CDK6 (Fig S10D). Therefore, the impact of CDK6 deletion on γ-globin expression depends on its kinase function.

Since RB1 is a canonical phosphorylation target of CDK6, we downregulated RB1 in primary human erythroid cells (Fig S10E) and demonstrated no impact of the latter on F-cell production (Fig S10F). Additionally, we showed that p27Kip1 overexpression resulted in a similar induction of F-cells, with and without RB1 downregulation (Fig S10F-G).

p27Kip1 increases F-cells in a mechanism likely independent of BCL11A and ZBTB7A

To determine if p27Kip1 mediates γ-globin expression by inhibiting either BCL11A45 or ZBTB7A46, we overexpressed CDKN1B-mTagBFP2 (or mTagBFP2) in HSPC-derived erythroid cells and transduced these cells with lentiviral particles expressing BCL11A (BCL11A-Puro) or ZBTB7A (ZBTB7A-Puro). We validated overexpression of BCL11A and ZBTB7A by RT-qPCR (Fig S9E). As expected, overexpression of BCL11A and ZBTB7A results in reduced frequency of F-cells in mTagBFP2 expressing control cells (Fig 6K). However, p27Kip1 overexpression resulted in comparable fold increases in F-cell percentage in control cells, cells overexpressing BCL11A, and cells overexpressing ZBTB7A (Fig 6K). These findings suggest that p27Kip1 overexpression results in increased F-cells in a mechanism likely independent of BCL11A and ZBTB7A.

Hydroxyurea may increase F-cells partially by CDK4/6 downregulation/inhibition

Hydroxyurea has a known effect on the cell cycle and induces γ-globin expression by a mechanism that is not completely understood22. To determine if hydroxyurea increases γ-globin production by directly or indirectly inhibiting CDK4/6, we treated HSPC-differentiated erythroid cells with hydroxyurea (or vehicle control) and transduced the erythroid cells with lentiviral particles expressing CDK4/6 (or control EV particles).

As expected, cells transduced with EV particles exhibited an increased proportion of F-cells following hydroxyurea treatment (Fig 6L), with the latter partially blunted by CDK4/6 overexpression (Fig 6L). These findings suggest that hydroxyurea may increase F-cells by a mechanism that depends partially on CDK4/6 downregulation and/or inhibition.

Discussion

The mechanism by which F-cells are formed is not well understood. In this report, our observation that slower cycling erythroid cells exhibit an increased proportion of F-cells led us to functionally test for a causal relationship between slowing the cell cycle speed (using CDKI overexpression) and γ-globin production. Here we show that p27Kip1 is the only CDKI that regulates γ-globin production and that the impact of p27Kip1 overexpression on γ-globin production is through its interaction with CDKs and/or cyclins, since i) expression of CDKN1B mutants that are unable to inhibit CDKs and/or cyclins did not result in a similar γ-globin induction and ii) pharmacologic inhibition and genetic downregulation of p27Kip1’s canonical targets, CDK4 and CDK6, also resulted in increased γ-globin production. Furthermore, using forced CDK4/6 expression in a rescue experiment, we showed that the increased γ-globin production mediated by p27Kip1 overexpression appears to be due to CDK4/6 downregulation. Consistent with these findings, we have shown that p27Kip1 overexpression results in macrocytosis, as has been reported with CDK4/6 inhibition, which was attributed to reduced numbers of erythroid cell divisions47.

Our observation that p27Kip1 is the only CDKI that induces γ-globin, despite all CDKIs resulting in reduced cell proliferation, suggests that slowing the cell cycle speed is not sufficient by itself to induce γ-globin expression. This finding can potentially be explained by recent studies demonstrating that CDKIs exhibit varying abilities to bind/inhibit various CDKs17,19,48. Furthermore, a recent proteomic analysis demonstrated that i) p27Kip1 and p19INK4d are the only CDKIs detected in HSPC-derived erythroblasts, ii) p27Kip1 protein is ~18-folds more abundant than p19INK4d in orthochromatic erythroblasts, and iii) that the proportion of p27Kip1 (but not p19INK4d) transcripts increases throughout erythroid differentiation35,36. These findings are consistent with a functional role of p27Kip1 in human erythropoiesis. Indeed, p27Kip1 was shown to promote cell cycle exit during the final stages of terminal erythroid maturation16. Our report demonstrates an additional, previously unrecognized role for p27Kip1 in γ-globin production.

We have shown that the impact of p27Kip1 overexpression on γ-globin production is mediated by CDK4/6 inhibition or modulated by CDK4/6 interaction. We also showed that p27Kip1 induces γ-globin production at the transcriptional level, resulting in increased γ-globin mRNA levels. Our data is consistent with two possibilities: i) CDK4/6 activity could repress a factor that promotes γ-globin expression, or ii) p27Kip1 that is not complexed with CDK4/6 interacts with a factor to promote γ-globin expression. In both cases, removing CDK4/6, or preventing p27Kip1 interaction with CDK4/6 complexes by palbociclib49, would promote γ-globin expression. Similarly, in both cases ectopic CDK4/6 expression would limit γ-globin expression resulting from p27Kip1 overexpression. Multiple mechanisms have been proposed for palbociclib function, and further work is needed to define the roles of p27Kip1 and palbociclib in γ-globin production49,50. Our findings that the increased γ-globin production mediated by CDK6 deletion was rescued by cDNA expression of wildtype CDK6 but not CDK6 K43M, suggests that the impact of CDK6 deletion on γ-globin expression depends on its kinase function.

The effect of p27Kip1 overexpression on γ-globin production seems to be insensitive to partial depletion of RB1, suggesting CDK6-mediated phosphorylation of non-canonical targets may be involved. However, the degree of RB1 downregulation achieved with the shRNA does not rule out a potential role for RB1 on γ-globin production. The effect of RB1 on cell cycle depends in part on the ratio of phospho-RB1 to total RB1, rather than the absolute level of RB1 expression. Additionally, other RB family members (p107 and p130) can be phosphorylated by CDK4/6 and may compensate for downregulation of RB1. Using overexpression rescue experiments, our findings suggest that p27Kip1 induces γ-globin production by a mechanism likely independent of BCL11A or ZBTB7A. While the overexpression level of BCL11A or ZBTB7A achieved in this experiment was not profound, it was sufficient to reduce γ-globin expression in control cells, suggesting that the modest overexpression of BCL11A or ZBTB7A attained is not the likely explanation of the lack of rescue.

A previous report demonstrated that F-cells exhibit similar transcriptomes and proteomes compared to non-F cells, suggesting that F-cell formation may be stochastic or regulated post-transcriptionally51,52. Our studies raise the possibility that slow cycling erythroid cells, due to increased p27Kip1 or reduced CDK4/6 function, have a higher predilection to become F-cells.

Previous work suggested that the structure and duration of the cell cycle regulate erythroid differentiation4. For example, differentiation of erythroleukemia cells was shown to require elongation of the G1 phase19, while entry and progression through S-phase was found to be required for activation of the terminal erythropoiesis program in murine fetal erythroblasts7,8. Our findings raise the possibility that elongation of G1 by p27Kip1 overexpression or CDK4/6 inhibition could result in preferential γ-globin (versus β-globin) transcription, through an undefined mechanism. Alternatively, p27Kip1 overexpression may result in increased or reduced expression of a gene(s) that induces or represses γ-globin, respectively. Future transcriptomic, proteomic, and phosphoproteomic studies in erythroid cells with and without p27Kip1 overexpression (or with and without CDK4/6 downregulation) may help identify how p27Kip1 regulates γ-globin production more precisely.

For several decades, hydroxyurea, a ribonucleotide reductase inhibitor that slows the cell cycle during S-phase, has been used clinically to treat SCD22,23. Hydroxyurea was proposed to induce γ-globin production by provoking stress erythropoiesis20,21. However, hydroxyurea does not appear to activate the HIF pathway41,51. Thus, stress erythropoiesis is unlikely to explain how hydroxyurea induces HbF. Despite additional mechanisms proposed53,54, the mechanism by which hydroxyurea increases HbF production remains unclear and recent work demonstrated no transcriptional differences between hydroxyurea-treated HbF and HbA-expressing cells51. Our work supports an additional mechanism by which hydroxyurea may contribute to γ-globin induction, based on the finding that CDK4/6 overexpression partially impairs the induction of γ-globin mediated by hydroxyurea. Seemingly paradoxically, hydroxyurea can lead to CDKI protein degradation, which for p27Kip1, may be via caspase cleavage55. However, effects of hydroxyurea on CDKIs are cell-type dependent, and hydroxyurea has been shown to increase the CDKI p21Cip1 in human erythroleukemic cells56. In addition, hydroxyurea treatment leading to arrest in the G1 phase of the cell cycle, as opposed to S-phase, does not trigger degradation of the CDKI p21Cip1, suggesting the timing of the hydroxyurea-induced cell cycle arrest during differentiation may have critical impacts on CDKI levels57. Hydroxyurea could impact HbF by altering p27Kip1 levels or may trigger CDK4/6 inhibition via degradation of a p27Kip1/Cyclin/CDK4 or CDK6 complex. With either model, future studies focused on investigating whether hydroxyurea and CDK4/6 inhibitors result in a synergistic induction in HbF would be of interest.

β-thalassemia and SCD pose substantial global health challenges. Although significant advances have been made with gene addition therapy and CRISPR/Cas9-based therapeutics, their accessibility remains limited in countries where most affected individuals reside1,2. Therefore, the development of innovative therapeutic approaches using compounds that are both effective and accessible is crucial. We found that palbociclib, an oral CDK4/6 inhibitor that is approved by the Food and Drug Administration for the treatment of a subset of breast cancer patients58, resulted in increased γ-globin production both in primary human erythroid cells and in the Townes SCD mouse model. These findings suggest that repurposing palbociclib (or other CDK4/6 inhibitors) may be beneficial for patients with β-thalassemia and SCD.

Supplementary Material

Supplement

Main points.

  • Overexpression of CDKN1B which encodes p27Kip1 leads to γ-globin induction via CDK4/6 inhibition.

  • CDK4/6 inhibitors increase HbF in human primary erythroid cells and in a murine SCD model and may thus be useful for β-hemoglobinopathies.

Acknowledgements

This work was supported by National Institutes of Health grants R01 HL148333 (R.Kh.), R01 HL157062 (R.Kh.), and U2C DK129445 (R.Kh.). G.B.C. was supported by the NIH F31- HL162544, NIH T32-GM007315, and the American Society of Hematology (MHGA). C.D. was supported by NIH T32-GM007315. G.My. was supported by National Institutes of Health grants U2C DK129445 and TL1DK136046. S.A.S. was supported by NIH research grant K08 DK127013. The author would like to thank Dr. Benjamin E. Tourdot for his insightful comments and input on the manuscript.

Footnotes

Competing Interests: The authors declare that they have no competing interests.

Data and materials availability:

All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials.

References

  • 1.Taher AT, Musallam KM, and Cappellini MD, beta-Thalassemias. N Engl J Med, 2021. 384(8): p. 727–743. [DOI] [PubMed] [Google Scholar]
  • 2.Kato GJ, et al. , Sickle cell disease. Nat Rev Dis Primers, 2018. 4: p. 18010. [DOI] [PubMed] [Google Scholar]
  • 3.Venkatesan V, et al. , Manipulation of Developmental Gamma-Globin Gene Expression: an Approach for Healing Hemoglobinopathies. Mol Cell Biol, 2020. 41(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Socolovsky M, Pas de deux: the coordinated coupling of erythroid differentiation with the cell cycle. Curr Opin Hematol, 2024. 31[1]: p. 96–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Lu YC, et al. , The Molecular Signature of Megakaryocyte-Erythroid Progenitors Reveals a Role for the Cell Cycle in Fate Specification. Cell Rep, 2018. 25(8): p. 2083–2093 e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Tusi BK, et al. , Population snapshots predict early haematopoietic and erythroid hierarchies. Nature, 2018. 555(7694): p. 54–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Pop R, et al. , A key commitment step in erythropoiesis is synchronized with the cell cycle clock through mutual inhibition between PU.1 and S-phase progression. PLoS Biol, 2010. 8(9). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Hwang Y, et al. , Global increase in replication fork speed during a p57(KIP2)-regulated erythroid cell fate switch. Sci Adv, 2017. 3(5): p. e1700298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Hidalgo D, et al. , EpoR stimulates rapid cycling and larger red cells during mouse and human erythropoiesis. Nat Commun, 2021. 12(1): p. 7334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Sankaran VG, et al. , Cyclin D3 coordinates the cell cycle during differentiation to regulate erythrocyte size and number. Genes Dev, 2012. 26(18): p. 2075–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Jayapal SR, et al. , Cyclin A2 regulates erythrocyte morphology and numbers. Cell Cycle, 2016. 15(22): p. 3070–3081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Humbert PO, et al. , E2F4 is essential for normal erythrocyte maturation and neonatal viability. Mol Cell, 2000. 6(2): p. 281–91. [DOI] [PubMed] [Google Scholar]
  • 13.Kinross KM, et al. , E2f4 regulates fetal erythropoiesis through the promotion of cellular proliferation. Blood, 2006. 108[1]: p. 886–95. [DOI] [PubMed] [Google Scholar]
  • 14.Clark AJ, Doyle KM, and Humbert PO, Cell-intrinsic requirement for pRb in erythropoiesis. Blood, 2004. 104(5): p. 1324–6. [DOI] [PubMed] [Google Scholar]
  • 15.Sankaran VG, Orkin SH, and Walkley CR, Rb intrinsically promotes erythropoiesis by coupling cell cycle exit with mitochondrial biogenesis. Genes Dev, 2008. 22(4): p. 463–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Gnanapragasam MN, et al. , EKLF/KLF1-regulated cell cycle exit is essential for erythroblast enucleation. Blood, 2016. 128(12): p. 1631–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Hsieh FF, et al. , Cell cycle exit during terminal erythroid differentiation is associated with accumulation of p27(Kip1) and inactivation of cdk2 kinase. Blood, 2000. 96(8): p. 2746–54. [PubMed] [Google Scholar]
  • 18.Minami R, et al. , p16(INK4a) induces differentiation and apoptosis in erythroid lineage cells. Exp Hematol, 2003. 31(5): p. 355–62. [DOI] [PubMed] [Google Scholar]
  • 19.Matushansky I, Radparvar F, and Skoultchi AI, Reprogramming leukemic cells to terminal differentiation by inhibiting specific cyclin-dependent kinases in G1. Proc Natl Acad Sci U S A, 2000. 97(26): p. 14317–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Stamatoyannopoulos G, et al. , On the induction of fetal hemoglobin in the adult; stress erythropoiesis, cell cycle-specific drugs, and recombinant erythropoietin. Prog Clin Biol Res, 1987. 251: p. 443–53. [PubMed] [Google Scholar]
  • 21.Stamatoyannopoulos G, Control of globin gene expression during development and erythroid differentiation. Exp Hematol, 2005. 33[1]: p. 259–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.McGann PT and Ware RE, Hydroxyurea therapy for sickle cell anemia. Expert Opin Drug Saf, 2015. 14(11): p. 1749–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Piel FB, Steinberg MH, and Rees DC, Sickle Cell Disease. N Engl J Med, 2017. 377[1]: p. 305. [DOI] [PubMed] [Google Scholar]
  • 24.Lin Z, et al. , The Endoplasmic Reticulum Cargo Receptor SURF4 Facilitates Efficient Erythropoietin Secretion. Mol Cell Biol, 2020. 40(23). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.King R, et al. , SEC23A rescues SEC23B-deficient congenital dyserythropoietic anemia type II. Sci Adv, 2021. 7(48): p. eabj5293. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Khoriaty R, et al. , SEC23B is required for pancreatic acinar cell function in adult mice. Mol Biol Cell, 2017. 28(15): p. 2146–2154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Khoriaty R, et al. , Absence of a red blood cell phenotype in mice with hematopoietic deficiency of SEC23B. Mol Cell Biol, 2014. 34(19): p. 3721–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Ryan TM, Ciavatta DJ, and Townes TM, Knockout-transgenic mouse model of sickle cell disease. Science, 1997. 278(5339): p. 873–6. [DOI] [PubMed] [Google Scholar]
  • 29.Nguyen J, et al. , Phenotypic Characterization the Townes Sickle Mice. Blood, 2014. 124(21): p. 4916–4916. [Google Scholar]
  • 30.Khoriaty R, et al. , Pancreatic SEC23B deficiency is sufficient to explain the perinatal lethality of germline SEC23B deficiency in mice. Sci Rep, 2016. 6: p. 27802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Khoriaty R, et al. , Functions of the COPII gene paralogs SEC23A and SEC23B are interchangeable in vivo. Proc Natl Acad Sci U S A, 2018. 115(33): p. E7748–E7757. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Kurita R, et al. , Establishment of immortalized human erythroid progenitor cell lines able to produce enucleated red blood cells. PLoS One, 2013. 8[1]: p. e59890. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Myers G, Friedman A, Yu L, et al. A genome-wide screen identifies genes required for erythroid differentiation. Nat Commun, 2025. 16: p. 3488. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Hu J, et al. , Isolation and functional characterization of human erythroblasts at distinct stages: implications for understanding of normal and disordered erythropoiesis in vivo. Blood, 2013. 121(16): p. 3246–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Ludwig LS, et al. , Transcriptional States and Chromatin Accessibility Underlying Human Erythropoiesis. Cell Rep, 2019. 27(11): p. 3228–3240 e7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Karayel O, et al. , Integrative proteomics reveals principles of dynamic phosphosignaling networks in human erythropoiesis. Mol Syst Biol, 2020. 16(12): p. e9813. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Sharma SS, et al. , A p27Kip1 mutant that does not inhibit CDK activity promotes centrosome amplification and micronucleation. Oncogene, 2012. 31(35): p. 3989–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Lim S. and Kaldis P, Cdks, cyclins and CKIs: roles beyond cell cycle regulation. Development, 2013. 140(15): p. 3079–93. [DOI] [PubMed] [Google Scholar]
  • 39.Razavipour SF, Harikumar KB, and Slingerland JM, p27 as a Transcriptional Regulator: New Roles in Development and Cancer. Cancer Res, 2020. 80(17): p. 3451–3458. [DOI] [PubMed] [Google Scholar]
  • 40.Vlach J, Hennecke S, and Amati B, Phosphorylation-dependent degradation of the cyclin-dependent kinase inhibitor p27. EMBO J, 1997. 16(17): p. 5334–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Feng R, Mayuranathan T, Huang P, Doerfler PA, et al. Activation of γ-globin expression by hypoxiainducible factor 1α. Nature, 2022. 610(7933): p. 783–790. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Bard H, Fouron JC, Gagnon C, et al. Hypoxemia and increased fetal hemoglobin synthesis. J Pediatr, 1994. 124(6): p. 941–943. [DOI] [PubMed] [Google Scholar]
  • 43.DeSimone J, Biel SI, Heller P, et al. Stimulation of fetal hemoglobin synthesis in baboons by hemolysis and hypoxia. Proc Natl Acad Sci U S A, 1978. 75(6): p. 2937–2940. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Hu MG, Deshpande A, Schlichting N, et al. CDK6 kinase activity is required for thymocyte development. Blood, 2011. 117(23): p. 6120–6131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Sankaran VG, et al. , Human fetal hemoglobin expression is regulated by the developmental stage-specific repressor BCL11A. Science, 2008. 322(5909): p. 1839–42. [DOI] [PubMed] [Google Scholar]
  • 46.Masuda T, et al. , Transcription factors LRF and BCL11A independently repress expression of fetal hemoglobin. Science, 2016. 351(6270): p. 285–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Malumbres M, Sotillo R, Santamaría D, et al. Mammalian cells cycle without the D-type cyclin-dependent kinases Cdk4 and Cdk6. Cell, 2004. 118(4): p. 493–504. [DOI] [PubMed] [Google Scholar]
  • 48.Matushansky I, Radparvar F, and Skoultchi AI, Manipulating the onset of cell cycle withdrawal in differentiated erythroid cells with cyclin-dependent kinases and inhibitors. Blood, 2000. 96(8): p. 2755–64. [PubMed] [Google Scholar]
  • 49.Guiley KZ, et al. , p27 allosterically activates cyclin-dependent kinase 4 and antagonizes palbociclib inhibition. Science, 2019. 366(6471). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Pennycook BR and Barr AR, Palbociclib-mediated cell cycle arrest can occur in the absence of the CDK inhibitors p21 and p27. Open Biol, 2021. 11(11): p. 210125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Khandros E, et al. , Understanding heterogeneity of fetal hemoglobin induction through comparative analysis of F and A erythroblasts. Blood, 2020. 135(22): p. 1957–1968. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Raj A. and van Oudenaarden A, Nature, nurture, or chance: stochastic gene expression and its consequences. Cell, 2008. 135(2): p. 216–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Wang M, et al. , Hydroxyurea exerts bi-modal dose-dependent effects on erythropoiesis in human cultured erythroid cells via distinct pathways. Br J Haematol, 2002. 119(4): p. 1098–105. [DOI] [PubMed] [Google Scholar]
  • 54.Cokic VP, et al. , Hydroxyurea induces fetal hemoglobin by the nitric oxide-dependent activation of soluble guanylyl cyclase. J Clin Invest, 2003. 111(2): p. 231–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Kramer OH, et al. , Histone deacetylase inhibitors and hydroxyurea modulate the cell cycle and cooperatively induce apoptosis. Oncogene, 2008. 27(6): p. 732–40. [DOI] [PubMed] [Google Scholar]
  • 56.Rodrigue CM, et al. , Resveratrol, a natural dietary phytoalexin, possesses similar properties to hydroxyurea towards erythroid differentiation. Br J Haematol, 2001. 113(2): p. 500–7. [DOI] [PubMed] [Google Scholar]
  • 57.Gottifredi V, et al. , Decreased p21 levels are required for efficient restart of DNA synthesis after S phase block. J Biol Chem, 2004. 279(7): p. 5802–10. [DOI] [PubMed] [Google Scholar]
  • 58.Lukasik P, et al. , Inhibitors of Cyclin-Dependent Kinases: Types and Their Mechanism of Action. Int J Mol Sci, 2021. 22(6). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement
NIHMS2179729-supplement-Supplement.pdf (3.1MB, pdf)

Data Availability Statement

All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials.

RESOURCES