The retinoblastoma (Rb), cyclin-dependent kinase (CDK), and CDK inhibitor genes regulate cell generation, and deregulation can produce increased cell growth and tumorigenesis. Polycythemia vera (PV) is a clonal myeloproliferative disease where the mechanism producing increased hematopoiesis is still unknown. To investigate possible defects in cell-cycle regulation in PV, the expression of Rb and CDK inhibitor gene messenger RNAs (mRNAs) in highly purified human erythroid colony-forming cells (ECFCs) was screened using an RNase protection assay (RPA) and 11 gene probes. It was found that RNA representing exon 2 of p16INK4a and p14ARF was enhanced by 2.8- to 15.9-fold in 11 patients with PV. No increase of exon 2 mRNA was evident in the T cells of patients with PV, or in the ECFCs and T cells from patients with secondary polycythemia. p27 also had elevated mRNA expression in PV ECFCs, but to a lesser degree. Because the INK4a/ARF locus encodes 2 tumor suppressors, p16INK4a and p14ARF with the same exon 2 sequence, the increased mRNA fragment could represent either one. To clarify this, mRNA representing the unique first exons of INK4a and ARF were analyzed by semiquantitative reverse transcription–polymerase chain reaction. This demonstrated that mRNAs from the first exons of both genes were increased in erythroid and granulocyte-macrophage cells and Western blot analysis showed that the INK4a protein (p16INK4a) was increased in PV ECFCs. Sequencing revealed no mutations of INK4a or ARF in 10 patients with PV. p16INK4a is an important negative cell-cycle regulator, but in contrast with a wide range of malignancies where inactivation of theINK4a gene is one of the most common carcinogenetic events, in PV p16 INK4a expression was dramatically increased without a significant change in ECFC cell cycle compared with normal ECFCs. It is quite likely that p16INK4a and p14ARF are not the pathogenetic cause of PV, but instead represent a cellular response to an abnormality of a downstream regulator of proliferation such as cyclin D, CDK4/CDK6, Rb, or E2F. Further work to delineate the function of these genes in PV is in progress.

Polycythemia vera (PV) is a member of the clonal myeloproliferative disorders, a spectrum of diseases resulting from the transformation of a pluripotent hematopoietic stem cell, and carrying a very high incidence of secondary acute leukemia.1-5 PV is characterized by trilineage marrow hyperplasia with increased production of red cells, granulocytes, and platelets, but it does not involve the immune system. Patients with PV have erythroid progenitors that develop in vitro without the addition of erythropoietin (EPO), displaying either an exquisite sensitivity to EPO or EPO-independent growth, and their blood EPO levels are generally quite low.6-8 

Past studies have shown that PV hematopoietic progenitors display hypersensitive responses to a variety of growth factors and cytokines. The burst-forming units-erythroid (BFU-E) are hypersensitive to interleukin-3 (IL-3), stem cell factor (SCF), granulocyte-macrophage colony-stimulating factor (GM-CSF), and insulinlike growth factor.9-14 Despite these abnormal responses, the number of receptors for the growth factors and their dissociation constants are normal.10-16 High levels of Bcl-x were identified in PV cells by one group of investigators and the investigators postulated that this might be responsible for increased survival of PV cells without EPO.17 Whereas a culture of normal erythroid progenitors with orthovanadate, an inhibitor of protein tyrosine phosphatases (PTPs), resulted in an increased number of erythroid colonies and enhanced protein tyrosine phosphorylation, little enhancement was evident with PV cells, suggesting that patients with PV may have an abnormal PTP activity which promotes increased cell proliferation.18 Further investigations revealed that the total PTP activity in PV erythroid cells was 3-fold higher than in normal cells.19 However, the increase in PTP may represent an effect of the disease rather than a cause. SHP-1 was completely characterized in PV cells and was found to be normal,19,20although one group provided evidence for its underexpression.21 Despite the development of much new knowledge on the control of erythropoiesis, the precise molecular defect that leads to enhanced hematopoiesis in PV is still not apparent. Unlike the situation with chronic myelogenous leukemia (CML), where the Philadelphia chromosome and an abnormal, specificBCR-ABL gene arising from a translocation have been identified, no pathognomonic chromosomal or gene abnormality has been found in PV.

Because the proliferation of normal cells is regulated by a combination of stimulatory and inhibitory factors that can respond to external signals in a coordinated manner, a permanent alteration within this regulatory system can lead to abnormal proliferation resulting in tumor formation.22-26 Inactivation of various cell proliferation–regulating proteins (such as the Rb, CDK, and CDK inhibitor proteins), due to gene deletions or mutations, results in enhanced cell generation.24 The INK4a/ARF locus provides important CDK inhibitors that include p16INK4a and p14ARF (p19ARF in the mouse).26-28p16INK4a inhibits CDK4 and CDK6 by directly blocking cyclin D–dependent kinase activity and, thereby, functions to suppress tumor cell growth.26,28 p16INK4a is inactivated by deletions, mutations, and methylation of a CpG island in many primary tumor types such as familial melanomas,29 esophageal carcinosarcomas,30 liver and lung carcinomas,31-33 and lymphomas and leukemias.34-37 Whereas p19ARF also inhibits cell proliferation, it is without direct inhibition of human CDKs and the process is p53-dependent as demonstrated by gene knockouts in mice.26,38 Other CDK inhibitors such as p27 may have a critical role in regulating entry into and exit from the mitotic cycle in response to extracellular signals.24 

We previously established a method to purify the early-stage erythroid progenitors (BFU-E) and grow late-stage erythroid colony-forming cells (ECFCs) from human peripheral blood39 and in past studies have repeatedly noticed that PV BFU-E generate many more ECFCs after 8 days of cell culture than normal BFU-E. Because PV is a proliferative disease where a principle problem is erythrocytosis, we have used highly purified erythroid progenitor cells from patients with PV and healthy donors to study the mechanism for increased cellular proliferation in this condition.

Generation of ECFCs and colony-forming units–granulocyte-macrophage

This method has been previously described.39 In brief, 400 mL of blood was obtained from healthy donors and from patients meeting the established criteria for PV or secondary polycythemia (SP)5,40 who signed consent forms approved by the Vanderbilt Committee for the Protection of Human Subjects and the Nashville Department of Veterans Affairs Research and Development Committee. The characteristics of our patients are listed in Table1. BFU-E were purified by sequential density gradient centrifugation; depletion of platelets and lymphocytes; and removal of adherent cells after overnight culture. A further negative selection and removal of contaminant cells with CD2, CD11b, CD16, and CD45 monoclonal antibodies were performed as previously described.39 41 The day-1 BFU-E were suspended in Iscoves modified Dulbecco medium (IMDM) containing 20% heat-inactivated fetal calf serum (FCS); 5% heat-inactivated, pooled, human AB serum; 1% deionized bovine serum albumin (BSA); 5 × 10−5 M 2-mercaptoethanol; 10 μg/mL insulin; 2 U/mL EPO; 50 U/mL IL-3; 100 ng/mL SCF; and streptomycin plus penicillin to generate ECFCs. After 7 days of culture the average purity of day-8 ECFCs was 90% or higher as measured by the plasma clot assay. In some experiments, the BFU-E were incubated for 5 to 18 days.

Table 1.

Patients with polycythemia

NumberSexAgePCVDiagnosis
PV1 59 47 polycythemia vera  
PV2 69 46 polycythemia vera 
PV3 52 44 polycythemia vera 
PV4 69 49 polycythemia vera 
PV5 68 50 polycythemia vera 
PV6 49 46 polycythemia vera 
PV7 72 45 polycythemia vera 
PV8 46 47 polycythemia vera 
PV9 61 46 polycythemia vera 
PV10 69 48 polycythemia vera 
PV11 74 50 polycythemia vera 
PV12 61 48 polycythemia vera 
SP1 49 54 hemoglobin with increased oxygen affinity 
SP2 48 57 carbon monoxide toxicity 
SP3 68 59 testosterone treatment for impotence 
SP4 60 53 carbon monoxide toxicity 
SP5 70 53 carbon monoxide toxicity 
SP6 57 52 carbon monoxide toxicity 
NumberSexAgePCVDiagnosis
PV1 59 47 polycythemia vera  
PV2 69 46 polycythemia vera 
PV3 52 44 polycythemia vera 
PV4 69 49 polycythemia vera 
PV5 68 50 polycythemia vera 
PV6 49 46 polycythemia vera 
PV7 72 45 polycythemia vera 
PV8 46 47 polycythemia vera 
PV9 61 46 polycythemia vera 
PV10 69 48 polycythemia vera 
PV11 74 50 polycythemia vera 
PV12 61 48 polycythemia vera 
SP1 49 54 hemoglobin with increased oxygen affinity 
SP2 48 57 carbon monoxide toxicity 
SP3 68 59 testosterone treatment for impotence 
SP4 60 53 carbon monoxide toxicity 
SP5 70 53 carbon monoxide toxicity 
SP6 57 52 carbon monoxide toxicity 

Patients with PV were treated only with phlebotomy except PV7, who received 1000 mg hydroxyurea per day. PV10 and PV11 received32P in 1997, but subsequently relapsed and were being phlebotomized. All patients with SP were treated only with phlebotomy.

PV indicates polycythemia vera; SP, secondary polycythemia.

Blood mononuclear cells were highly enriched in colony-forming units–granulocyte-macrophage (CFU-GM) using the method for BFU-E except that negative selection was performed without CD45 as previously described,10 but with the addition of CD20. The average purity of CFU-GM was 40 ± 6% by the standard colony-forming assay.10 Culture of these cells was performed in a modified liquid medium for 7 days without EPO.10 

Preparation of proliferating T cells

T cells were isolated from the blood of healthy individuals and patients after sheep erythrocyte rosetting during the purification of BFU-E.41 The sheep erythrocytes were lysed in sterile water and the T cells were incubated at 37°C in RPMI 1640 medium including 10% FCS, 40 μg/mL streptomycin, 500 U/mL penicillin, and 3 μg/mL phytohemagglutinin (PHA) for 24 hours. The cells were then washed and resuspended in the same medium without PHA. IL-2 at a concentration of 30 U/mL was added for a 72-hour incubation at 37°C, and then the medium was replaced with fresh IL-2 medium before the cells were collected at 120 hours. Cell-cycle analysis by flow cytometry showed that at least 30% of the T cells were in the S or M phases of the cell cycle (data not shown).

RNA extraction and ribonuclease protection assay

Total RNA was prepared from day-8 ECFCs or T cells using Ultraspec RNAzol (Biotecx Laboratories, Houston, TX) and ribonuclease protection assays (RPAs) were performed using the MAXIscript and RPA II Ribonuclease Protection Assay Kit (Ambion, Austin, TX) according to the manufacturers' protocols. The hCC-2 Multi-Probe Template Set and individual, custom probes for expression of p130, p107, p53, p57, p27, p21, p19, p18, p16, p14/15, and Rb RNA were purchased from Pharmingen (San Diego, CA). The protected transcripts were separated on denaturing polyacrylamide gels, quantified with a laser scanning densitometer, and normalized to the amount of total RNA present in the lane by comparison with the housekeeping gene signals. Gels were developed at varying times so that we could scan the L32 and GADPH housekeeping gene RNAs as internal controls, to correct for variations in loading.

INK4a and ARF sequence analysis

Genomic DNAs from day-8 ECFCs from healthy patients and patients with PV were prepared using the QIAamp Blood Kit (Qiagen, Valencia, CA) according to the manufacturer's instructions. Four hundred nanograms genomic DNA template was used in each reaction. The INK4a/ARF DNA has a high G:C ratio and the specific fragments could not be successfully amplified with the usual PCR kit, so an Advantage-GC complementary DNA (cDNA) PCR kit was purchased from Clontech (Palo Alto, CA). The reaction solution contained GC cDNA PCR buffer, GC melt, 10 mM each of dATP, dTTP, dCTP, dGTP, DNA polymerase, and specific primers for INK4a/ARF exon 1α, exon 1β, exon 2, and exon 3. The intron oligonucleotide primers were synthesized by Gibco BRL, Life Technologies (Rockville, MD) or Research Genetics (Huntsville, AL) and the sequences of the primers are listed in Table2. Forty cycles were performed and the conditions and cycles were as follows: 5 minutes at 95°C; then 10 cycles with denaturation at 95°C for 1 minute; annealing at 68°C for 1 minute and extension at 72°C for 1 minute; then 10 cycles at 95°C for 1 minute, 64°C for 1 minute, and 72°C for 1 minute; then 20 cycles at 95°C for 1 minute, 60°C for 1 minute, and 72°C for 1 minute; and final elongation at 72°C for 10 minutes. The amplified PCR products were electrophoresed in 2% agarose gels to confirm expected single bands. The PCR products were then cloned into the PCR 2.1 TA cloning vector using the Original TA Cloning Kit (Invitrogen, San Diego, CA) and introduced into competent αH5 cells. The plasmid DNAs were purified using the Qiagen QIAprep Kit. Restriction analyses were performed to determine the presence of the insert and sequence analyses were performed at the Vanderbilt Sequencing Center, Nashville, TN.

Table 2.

Polymerase chain reaction (PCR) and reverse transcription–PCR primer sequences

Primers5′ to 3′ sequencesAmplification bp
PCR   
 Exon 1α  340 
  Forward GAA GAA AGA GGA GGG GCT G  
  Reverse GCG CTA CCT GAT TCC AAT TC  
 Exon 2  394  
  Forward TTC CTT TCC GTC ATG CCG G  
  Reverse GTA CAA ATT CTC AGA TCA TCA GTC GTC  
 Exon 3  323 
  Forward GAC CTG GAG CGC TTG AGC GG  
  Reverse GTG GCC CTG TAG GAC CTT CGG  
 Exon 1β  440  
  Forward TCC CAG TCT GAC GTT AAG GGG  
  Reverse GTC TAA GTC GTT GTA ACC CG  
RT-PCR   
 INK4a (exon 1α to exon 3)  494  
  Forward ATG GAG CCT TCG GCT GAC TGG C  
  Reverse GAC TGA TGA TCT AAG TTT CCC GAG G  
 ARF (exon 1β to exon 2)  571 
  Forward GAG TGG CGC TGC TCA CCT CTG G  
  Reverse GGG ACC TTC CGC GGC ATC  
 INK4a  142  
  Forward ATG GAG CCT TCG GCT GAC TGG C  
  Reverse CTG CCC ATC ATC ATG ACC TGG A  
 ARF  212 
  Forward CAT GGT GCG CAG GTT CTT GG  
  Reverse same as INK4a  
Primers5′ to 3′ sequencesAmplification bp
PCR   
 Exon 1α  340 
  Forward GAA GAA AGA GGA GGG GCT G  
  Reverse GCG CTA CCT GAT TCC AAT TC  
 Exon 2  394  
  Forward TTC CTT TCC GTC ATG CCG G  
  Reverse GTA CAA ATT CTC AGA TCA TCA GTC GTC  
 Exon 3  323 
  Forward GAC CTG GAG CGC TTG AGC GG  
  Reverse GTG GCC CTG TAG GAC CTT CGG  
 Exon 1β  440  
  Forward TCC CAG TCT GAC GTT AAG GGG  
  Reverse GTC TAA GTC GTT GTA ACC CG  
RT-PCR   
 INK4a (exon 1α to exon 3)  494  
  Forward ATG GAG CCT TCG GCT GAC TGG C  
  Reverse GAC TGA TGA TCT AAG TTT CCC GAG G  
 ARF (exon 1β to exon 2)  571 
  Forward GAG TGG CGC TGC TCA CCT CTG G  
  Reverse GGG ACC TTC CGC GGC ATC  
 INK4a  142  
  Forward ATG GAG CCT TCG GCT GAC TGG C  
  Reverse CTG CCC ATC ATC ATG ACC TGG A  
 ARF  212 
  Forward CAT GGT GCG CAG GTT CTT GG  
  Reverse same as INK4a  

RT indicates reverse transcription; bp, base pair.

RT-PCR and semiquantitative RT-PCR

First strand cDNA was synthesized from day-8 normal and PV ECFC RNA using the SuperScript First-Strand Synthesis System for RT-PCR (Gibco BRL). One microgram total RNA was incubated with the components in the kit following the manufacturer's protocol. The final volume of the reaction was 20 μl. Two micrograms of the cDNA and the Advantage-GC cDNA PCR kit (described above) were used for PCR. The specific primers designed to amplify INK4a and ARF are listed in Table2. First we designed a pair of primers to amplify a 494-bp product, from exon 1α through exons 2 and 3 for INK4a, and a pair of primers to generate a 571-bp product for ARF from exon 1β through exon 2. The conditions for INK4a PCR were as follows: 95°C for 5 minutes; followed by 10 cycles at 95°C for 1 minute, 64°C for 1 minute, and 72°C for 1 minute; then 10 cycles at 95°C for 1 minute, 60°C for 1 minute, and 72°C for 1 minute; then 20 cycles at 95°C for 1 minute, 56°C for 1 minute, and 72°C for 1 minute; then the final step at 72°C for 10 minutes. The conditions for ARF PCR were: 95°C for 5 minutes; then 10 cycles at 95°C for 1 minute, 66°C for 1 minute and 72°C for 1 minute; 10 cycles at 95°C for 1 minute, 64°C for 1 minute and 72°C for 1 minute; then 20 cycles at 95°C for 1 minute, 62°C for 1 minute, and 72°C for 1 minute, followed by the final step at 72°C for 10 minutes. For semiquantitative RT-PCR, the cDNAs were diluted 1:4, 1:16, and 1:64 using the PCR reaction solution and PCR was performed using the same method described above. Table 2 also shows the sequence of shorter primers for amplifying both INK4a (142 bp), from exon 1α through part of exon 2, and ARF (212 bp), from exon 1β through part of exon 2. The PCR conditions were the same as those stated above for the ARF PCR. The products were electrophoresed in 2% agarose gels and quantified with a laser scanning densitometer.

Western blot analysis

Cell extracts were prepared by lysing 107 cells in 100 μl RIPA lysis buffer containing 1% Triton X-100, 20 mM Tris-HC1, ph 7.5, 10% glycerol, 140 mM NaCl, 100 mM sodium fluoride, 10 mM EDTA, 2 mM vanadate, 0.2 mM phenylmethylsulfonyl fluoride, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), and 0.15 U/mL aprotinin at 4°C. Insoluble materials were removed by centrifugation for 20 minutes at 14 000g and 4°C. The samples were quantitated using a Bio-rad protein assay kit II (Bio-rad, Hercules, CA) and were boiled for 5 minutes in an SDS sample buffer before 15% SDS-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were transferred to nitrocellulose after SDS-PAGE. The blots were incubated for 2 hours at room temperature in Tris-buffered saline with 0.05% Tween-20 (TBST) and 5% nonfat milk and then with anti-p16INK4a overnight at 4°C. The antibodies to p16INK4a were purchased from Santa Cruz (SC-759) (Santa Cruz Biotechnology, Santa Cruz, CA) and Pharmingen and the antibody to p14ARF (clone 14P02) was purchased from Neomarkers (Fremont, CA).

Cell-cycle analysis

Day-6 to day-19 cells (1-2 × 106) were washed and the cell pellets were resuspended in propidium iodide solution (0.05 mg/mL in 0.1% sodium citrate) and RNase solution (0.1 mg/mL).42 The cell samples generally were examined within 30 minutes of staining at 4°C by the Vanderbilt Flow Cytometry Lab, but some samples were fixed in 70% ethanol for analysis at a later time.

Screen of cell proliferation regulatory gene expression in PV ECFCs

Highly purified day-8 ECFCs were generated from the blood BFU-E of 6 patients with PV and 6 healthy subjects and RNA was extracted. Twenty microgram samples were analyzed for the presence of messenger RNA (mRNA) transcripts using the hCC-2 human cell-cycle Multi-Probe Template Set. This contains templates that can be used for T7 polymerase-directed synthesis of high specific activity [32P]-labeled antisense RNA probes, which can hybridize with target human mRNAs for p130, Rb, p107, p53, p57, p27, p21, p19, p18, p16 exon 2, p14/15, and the housekeeping genes L32 andGAPDH as internal controls. RPAs were performed, and protected transcripts were separated on denaturing polyacrylamide gels and quantified by autoradiography (Figure1). The unprotected template set probes are shown on the left for comparison with the RNase-protected probes following hybridization with normal and PV total RNAs. All samples from the 6 patients with PV showed a well demarcated band with a strikingly greater density than that of the healthy controls. However, because the set contained multiple probes, it was difficult to be sure which specific gene was expressed very highly in these patients, so we next obtained all of the above individual probes from Pharmingen for further studies.

Fig. 1.

Abnormally high gene mRNA expression in PV erythroid progenitors.

Highly purified day-8 ECFCs were generated from 6 healthy donors and 6 patients with PV. Total RNAs were isolated and 20 μg of RNA were analyzed for the presence of RNA transcripts for p130, Rb, p107, p53, p57, p21, p19, p18, p16, p14/15 related to cell proliferation, using the hCC-2 Multi-Probe Template Set (Pharmingen). L32 andGAPDH were included as internal controls. RNase protection assays (RPAs) were performed with the MAXIscript and RPA II Ribonuclease Protection Assay Kits (Ambion) according to the manufacturer's protocols. Protected transcripts were separated by denaturing polyacrylamide gels and quantified by autoradiography. The band with a strikingly higher expression is indicated by an arrow.

Fig. 1.

Abnormally high gene mRNA expression in PV erythroid progenitors.

Highly purified day-8 ECFCs were generated from 6 healthy donors and 6 patients with PV. Total RNAs were isolated and 20 μg of RNA were analyzed for the presence of RNA transcripts for p130, Rb, p107, p53, p57, p21, p19, p18, p16, p14/15 related to cell proliferation, using the hCC-2 Multi-Probe Template Set (Pharmingen). L32 andGAPDH were included as internal controls. RNase protection assays (RPAs) were performed with the MAXIscript and RPA II Ribonuclease Protection Assay Kits (Ambion) according to the manufacturer's protocols. Protected transcripts were separated by denaturing polyacrylamide gels and quantified by autoradiography. The band with a strikingly higher expression is indicated by an arrow.

Close modal

p16 exon 2 gene expression is much greater in PV ECFs

Eleven individual gene probes, including all of those of the above set with internal controls L32 and GAPDH, were used for separate RPAs to measure mRNA expression in PV day-8 ECFCs and normal day-8 ECFCs. No differences were evident with most of the above individual gene probes between normal and PV samples, but the p16 mRNA transcript was significantly enhanced in all 11 studied patients with PV compared with 9 healthy donors. Figure2 (top panel) shows the results of RPAs for p16 exon 2 mRNA expression in ECFCs from 4 healthy donors and 4 patients with PV. Table 3 summarizes these results and the data show that p16 exon 2 transcripts from the 11 patients with PV were quantitatively increased by 2.8- to 15.9-fold as measured by densitometry. We also measured p16 exon 2 mRNA expression in day-10 ECFCs from 3 healthy patients and 4 patients with PV to determine if this increase was limited to a defined stage of erythroid maturation, and we found that no significant difference in p16 mRNA expression was evident, either in normal or PV ECFCs at this time compared with day-8 cells (data not shown).

Fig. 2.

Enhanced amount of mRNA transcripts for p16 exon 2 and p27 in the ECFCs of 4 patients with PV.

Experiments performed as in Figure 1, but with only p16, L32, and GAPDH gene probes or p27 and L32 and GAPDH.

Fig. 2.

Enhanced amount of mRNA transcripts for p16 exon 2 and p27 in the ECFCs of 4 patients with PV.

Experiments performed as in Figure 1, but with only p16, L32, and GAPDH gene probes or p27 and L32 and GAPDH.

Close modal
Table 3.

Expression of protected p16 exon 2 fragment in polycythemia vera and normal erythroid colony-forming cells

Exp no.Normal ECFCsPV ECFCsp16 density, healthy donorsp16 density, patients with PV% increase in PV
N7  1 073   
  PV3  17 044 1 590 
  PV6  5 863 550 
 N9  1 437   
  PV4  5 834 410 
  PV5  15 275 1 063 
N11  731   
  PV1  6 085 845 
N2  630   
  PV2  2 668 283 
N13  3 319   
  PV8  16 173 387 
N12  2 290   
  PV9  8 300 362 
N8  1 225   
  PV7  3 478 284 
N14  1 280   
  PV10  3 625 283 
 N15  651   
  PV11  3 707 439 
Exp no.Normal ECFCsPV ECFCsp16 density, healthy donorsp16 density, patients with PV% increase in PV
N7  1 073   
  PV3  17 044 1 590 
  PV6  5 863 550 
 N9  1 437   
  PV4  5 834 410 
  PV5  15 275 1 063 
N11  731   
  PV1  6 085 845 
N2  630   
  PV2  2 668 283 
N13  3 319   
  PV8  16 173 387 
N12  2 290   
  PV9  8 300 362 
N8  1 225   
  PV7  3 478 284 
N14  1 280   
  PV10  3 625 283 
 N15  651   
  PV11  3 707 439 

The density of p16 exon 2 assayed by RNase protection assay was quantified with a laser scanning densitometer and normalized to the amount of the total RNA present in the lane by comparison with the housekeeping gene mRNA. The percentage increase in patients with PV was obtained using the density in PV divided by the normal control density in the same gel so that all conditions were the same.

ECFCs indicates erythroid colony-forming cells; PV, polycythemia vera.

p27 gene expression is also increased in PV ECFCs, but is less than p16 expression

The amount of p27 mRNA was delineated in the same 4 normal and 4 PV samples by RPA and the results are shown in Figure 2 (bottom panel). The percentage increases of the p27 gene transcripts in comparison with the p16 exon 2 transcripts, in the same healthy and patient samples, are listed in Table 4. p27 mRNA expression also was enhanced in PV ECFCs, but the degree of increase was much lower than that seen with the p16 exon 2 mRNA.

Table 4.

Increase of p16 exon 2 and p27 messenger RNA in polycythemia vera

PV samplep16, % increase in PVp27, % increase in PV
PV3 1590 215  
PV6 550 304 
PV4 410 209  
PV5 1063 402 
PV samplep16, % increase in PVp27, % increase in PV
PV3 1590 215  
PV6 550 304 
PV4 410 209  
PV5 1063 402 

PV indicates polycythemia vera.

Lack of enhanced p16 exon 2 gene expression in patients with SP

Because p16 mRNA expression is greater than p27 gene expression, we focused our study on p16. PV is a clonal disease with increased red cell production and the increased p16 expression could simply be a manifestation of increased erythropoiesis. SP manifests increased erythropoiesis due to increased EPO production and we obtained ECFCs from the 6 patients with SP listed in Table 1. Figure3 shows a comparison of the amount of the p16 exon 2 transcript from 3 normal, 3 PV, and 3 SP ECFCs and it demonstrates that no increase was present in SP compared with PV. Table5 summarizes 5 experiments comparing the density of the p16 exon 2 transcripts in normal and SP ECFCs and it shows that no difference was present, indicating that the abnormality of p16 expression in PV does not appear to be related to increased erythropoiesis per se.

Fig. 3.

Enhanced amount of mRNA transcript for p16 (exon 2) in PV ECFCs is not present in secondary polycythemia.

Experiments performed as in Figure 1, but with only p16, L32, and GAPDH gene probes. SP indicates secondary polycythemia.

Fig. 3.

Enhanced amount of mRNA transcript for p16 (exon 2) in PV ECFCs is not present in secondary polycythemia.

Experiments performed as in Figure 1, but with only p16, L32, and GAPDH gene probes. SP indicates secondary polycythemia.

Close modal
Table 5.

Expression of protected p16 exon 2 fragment in secondary polycythemia and normal erythroid colony-forming cells

Experiment no.Normal ECFCsSP ECFCsp16 density, healthy donorsp16 density, patients with SP
N7  654  
  SP1  412 
  SP6  399 
N13  3319  
  SP5  4292 
N11  2290  
  SP2  1645 
N9  1225  
  SP4  533 
N12  1280  
  SP3  1036 
Experiment no.Normal ECFCsSP ECFCsp16 density, healthy donorsp16 density, patients with SP
N7  654  
  SP1  412 
  SP6  399 
N13  3319  
  SP5  4292 
N11  2290  
  SP2  1645 
N9  1225  
  SP4  533 
N12  1280  
  SP3  1036 

RNase protection assay experiments performed and measurements made as in Table 3. For abbreviations, see Tables 1 and 3.

Lack of enhanced p16 exon 2 gene expression in activated T cells of patients with PV

To determine if the increased p16 expression was a general manifestation of PV, or was restricted to the myeloproliferative clone, we analyzed activated PV T cells because they are not involved in the clonal PV disease. Blood T cells from healthy, PV, and SP donors were isolated and stimulated with phytohemagglutinin followed by IL-2 in order to provide a state of cell proliferation similar to that of the ECFCs. These activated T cells were then analyzed for p16 mRNA expression. We compared the p16 mRNA levels from activated T cells and day-8 ECFCs of the same patient with PV as well as from activated T cells of a second patient with PV and an SP and healthy donor (Figure 4). Activated PV T cells did not demonstrate enhanced expression of p16 exon 2 mRNA. Thus the PV abnormality does not reside in noninvolved tissues, but instead appears to be related to the clonal hematopoietic disease.

Fig. 4.

Enhanced expression of mRNA transcript for p16 (exon 2) is not present in activated PV T cells compared with PV ECFCs (left panel) and compared with activated T cells of a healthy or SP donor (right panel).

T cells from the BFU-E purification process were incubated for 24 hours with 3 μg/mL PHA. Cells were then washed and incubated for 5 days with 30 U/mL IL-2. RNA was then isolated and experiments were performed as in Figure 1, but with only p16 (exon 2), L32, and GAPDH gene probes. Activated PV T cells are compared with PV ECFCs (left panel) and with activated T cells of a healthy or SP donor (right panel).

Fig. 4.

Enhanced expression of mRNA transcript for p16 (exon 2) is not present in activated PV T cells compared with PV ECFCs (left panel) and compared with activated T cells of a healthy or SP donor (right panel).

T cells from the BFU-E purification process were incubated for 24 hours with 3 μg/mL PHA. Cells were then washed and incubated for 5 days with 30 U/mL IL-2. RNA was then isolated and experiments were performed as in Figure 1, but with only p16 (exon 2), L32, and GAPDH gene probes. Activated PV T cells are compared with PV ECFCs (left panel) and with activated T cells of a healthy or SP donor (right panel).

Close modal

RT-PCR demonstrates that both the p16INK4a and p14ARF transcripts are enhanced in PV ECFCs as well as PV myeloid cells

p16 is encoded in the INK4a/ARF locus at chromosome 9p21 and described as p16INK4a.26,28 A second transcript, p14ARF, has been identified at the same locus due to an unprecedented use of an alternative reading frame (ARF) and it also provides potent regulation of cell proliferation.26,28,38,43 The 2 transcripts (α for INK4a and β for ARF) arise from different promoters and have different first exons. They share exons 2 and 3 but, because of distinct reading frames in exon 2, no amino acid identity exists between the 2 products.26 38 Because the probe for the p16 transcript provided by Pharmingen was generated from exon 2 and because p16INK4a and p14ARF share this exon 2 sequence, the protected fragment that we detected could be a part of either one. To clarify this, we performed an analysis of p16INK4a and p14ARF mRNA expression by RT-PCR. Ten normal and 6 PV samples were studied. No amplification of p16INK4a was seen in any of the normal samples, but was clearly present in all PV samples except the sample from patient 4 (Figure5). The expression of p14ARFmRNA was equally strong in both the normal and PV ECFCs and was much greater than p16INK4a expression in both the normal and PV cells. However, because the amplified fragments were very long and the mRNA was GC rich, p16INK4a expression could not be confirmed in the normal ECFCs. Also, because the amplified p14ARF product was very strong in all samples, it needed to be quantified. Therefore, we designed another pair of primers for p16INK4a and p14ARF amplification with specific forward primers, but only a single reverse primer, for each, because both p16INK4a and p14ARF share the same exon 2 sequence. The length of the RT-PCR product for p16INK4a was 142 bp and covered almost the entire exon 1α and part of exon 2, whereas the length of p14ARF was 212 bp and covered almost the entire exon 1β and a part of exon 2. The result of a semiquantitative assay from a normal ECFC extract and a PV ECFC extract is shown in Figure 6 and demonstrates that both p16INK4a and p14ARF had enhanced expression in PV ECFCs compared with the normal control cells. This same experiment was repeated with 3 different normal and PV collections of ECFCs and showed the same results. Similar results were evident when purified CFU-GM were cultured for 7 days and analysis was performed in the same way (Figure 7).

Fig. 5.

Analysis of p16INK4a and p14ARFmRNA expression in normal and PV ECFCs by RT-PCR.

Information concerning primers and length of products is shown in Table2. The markers show 1000, 750, 500, 300, 150, and 50 bp. The length of p16INK4a is 494 bp and the length of p14ARF is 571 bp.

Fig. 5.

Analysis of p16INK4a and p14ARFmRNA expression in normal and PV ECFCs by RT-PCR.

Information concerning primers and length of products is shown in Table2. The markers show 1000, 750, 500, 300, 150, and 50 bp. The length of p16INK4a is 494 bp and the length of p14ARF is 571 bp.

Close modal
Fig. 6.

Semiquantitative assay of p16INK4a and p14ARF mRNA transcripts by RT-PCR in normal and PV ECFCs.

The sequence of primers is shown in Table 2. The length of the p16INK4a amplification product is 142 bp and the length of p14ARF is 212 bp. The marker sizes on the top panel are 750, 500, 300, 150, and 50 bp whereas the markers on the bottom panel are 1000, 750, 500, 300, and 150 bp. The amount of original total RNA used for the RT-PCR was 80 ng, and 20 ng, 5 ng and 1.25 ng, respectively, were used for the dilutions of 1:4 to 1:64.

Fig. 6.

Semiquantitative assay of p16INK4a and p14ARF mRNA transcripts by RT-PCR in normal and PV ECFCs.

The sequence of primers is shown in Table 2. The length of the p16INK4a amplification product is 142 bp and the length of p14ARF is 212 bp. The marker sizes on the top panel are 750, 500, 300, 150, and 50 bp whereas the markers on the bottom panel are 1000, 750, 500, 300, and 150 bp. The amount of original total RNA used for the RT-PCR was 80 ng, and 20 ng, 5 ng and 1.25 ng, respectively, were used for the dilutions of 1:4 to 1:64.

Close modal
Fig. 7.

Semiquantitative assay of p16INK4a and p14ARF mRNA transcripts by RT-PCR in normal and PV granulocyte-macrophage cells generated from blood CFU-GM.

The method is the same as in Figure 6.

Fig. 7.

Semiquantitative assay of p16INK4a and p14ARF mRNA transcripts by RT-PCR in normal and PV granulocyte-macrophage cells generated from blood CFU-GM.

The method is the same as in Figure 6.

Close modal

INK4a and ARF mutation analysis

Because a malfunction of p16 could lead to abnormal feedback and increased expression, each of the 3 exons 1α, 2, and 3 of INK4a and the alternatively spliced exon 1β of ARF were amplified by PCR from genomic DNA of healthy patients and patients with PV using primers complementary to sequences flanking each exon. Sequence analysis for each exon was performed, but no mutation was identified in 10 patients with PV.

Quantitation of INK4a protein (p16INK4a) in PV ECFCs

Western blot analysis of p16INK4a was performed on normal and PV ECFC protein. p16INK4a protein was not detectable in 6 normal ECFCs, but was detectable and increased in 4 of 6 PV ECFCs (Figure 8). No stable, reliable results were obtained with the only antihuman ARF protein antibody that was available to us.

Fig. 8.

INK4a protein (p16INK4a) expression in normal and PV ECFCs.

Two separate Western blots are shown in which PV3 and PV11 have very high levels of p16INK4a protein; PV1 and PV8 have lower levels. p16INK4a was virtually undetectable in all healthy ECFC protein extracts.

Fig. 8.

INK4a protein (p16INK4a) expression in normal and PV ECFCs.

Two separate Western blots are shown in which PV3 and PV11 have very high levels of p16INK4a protein; PV1 and PV8 have lower levels. p16INK4a was virtually undetectable in all healthy ECFC protein extracts.

Close modal

Normal and PV ECFC cell-cycle analysis

Cell-cycle analyses on normal and PV day-8 ECFCs were performed. In 5 normal samples the percentage of ECFCs in the G0/G1 stage was 42% ± 12%, S phase was 49% ± 9%, and the G2/M phase was 9% ± 4%. In 4 PV ECFC collections the percentage of cells in the G0/G1 phase was 41% ± 5%, S phase was 53% ± 6%, and the G2/M phase was 6% ± 3%. The percentages of PV and normal cells in the S/G2/M phases were not significantly different. The cell-cycle studies also were carried out on 3 normal and 3 PV ECFC samples in different stages of maturation: day-6, day-8, day-10, day-13, and day-19 cells. The average percentage of S/G2/M stage cells on day 6, day 8, day 10, day 13, and day 19 from healthy individuals was 48%, 67%, 50%, 39%, and 9%, respectively, whereas in PV ECFCs it was 54%, 64%, 53%, 42% and 11%, respectively, which were not statistically significant changes.

The INK4a/ARF locus on chromosome 9 is one of the sites mutated most frequently in human cancer.44 We have demonstrated in 11 different PV ECFCs, compared with normal ECFCs, a marked increase in the expression of p16INK4a, a unique inhibitor of cell proliferation and a tumor suppressor. The mRNA also was increased in highly purified PV myeloid cells, but enhanced expression was not present in PV T cells and also was not present in SP ECFCs. Thus the abnormal expression appears to be limited to cells of the myeloproliferative clone of PV, and it is not due to a mutation of the structural components of the gene, which might have decreased the gene's function, as sequence analysis of exons 1α, 2, and 3 in 10 patients with PV showed no mutations of INK4a. In addition, PV ECFCs were shown to have increased expression of p14ARF and p27 mRNA. This is different from the observations made with chronic myeloid leukemia cells where p27 mRNA levels are normal despite relocation of p27 to the cell cytoplasm and inactivation of the protein,45 and is yet another difference of PV from other myeloproliferative diseases.

In the normal cell cycle, control of progression from G1 to S phase is executed predominantly by the gatekeeper Rb protein. In G1, hypophosphorylated Rb binds to the E2F transcription factor, which results in transcriptional repression of E2F-responsive genes and blocks G1/S progression.25,26 In response to growth-promoting signals, a complex of cyclin D and CDK4/6 induces Rb phosphorylation with release of Rb from E2F allowing E2F to transactivate many genes important for mitosis. p16INK4ainhibits the kinase activity of CDK4/6 by changing the conformation of the cyclin-binding site, preventing ATP binding, and thereby producing hypophosphorylation of Rb, which in turn decreases the expression of the E2F-dependent genes.26 Thus p16INK4a is able to block passage from G1 into S and is a bona fide tumor suppressor. Ectopic expression of p16INK4a in human cells induces senescence, which can occur independently of p53.26 

Although p14ARF also inhibits the cell cycle, it is without direct inhibition of known CDKs and its effect is p53-dependent as demonstrated by gene knockouts in mice.26,38p14ARF binds to the C-terminal region of MDM2 through its exon 1β-encoding N-terminal domain and promotes the rapid degradation of MDM2, or sequesters MDM2 (sometimes called HDM2 for the human protein) into the nucleolus thereby preventing negative feedback regulation (or degradation) of p53 by MDM2.46,47 This permits p53-induced growth arrest or apoptosis. An ARF mutant that binds MDM2, but does not move it to the nucleolus, is unable to induce p53-dependent cell-cycle arrest.46 Most mutations of exon 2 that disable p16INK4a formation do not decrease the activity of p14ARF suggesting that exon 1β contributes most of the domains necessary for p14ARFfunction.26 Several mitogenic stimuli, E1A, myc, oncogenic ras, V-abl, and E2F-1 upregulate ARF leading to p53 stabilization. Of interest is the fact that mice lacking exons 2 and 3 showed extramedullary hematopoiesis in the absence of anemia.26 

Our results indicate that unlike several malignant diseases, PV has no mutation of p16INK4a or p14ARF. In contrast, both genes are expressed in an abnormally high amount in PV ECFCs and this is quite paradoxical, as PV is a disease with increased proliferation. We believe that the large increase in p16INK4a expression in a proliferative disease is most likely due to an abnormality of a downstream protein with which this protein interacts, thereby creating a cellular reaction or feedback for an elevated p16INK4a concentration, rather than an intrinsic abnormality of the p16INK4a promoter, as a primary defect of the latter would not account for the increased cell generation in the disease. Because p16INK4a inhibits cell generation by binding CDK4/6, and because our results demonstrate a p16INK4a increase in PV, it is possible that a mutation of CDK4/6 is present in PV which decreases its binding with p16INK4a and may lead to p16INK4a accumulation. A mutation in familial melanoma, arginine-to-cysteine exchange at residue 24, has been reported, which has a specific effect on the p16INK4a binding domain of CDK4/6, but has no effect on its ability to bind cyclin D and form a functional kinase.48 49 This R24C mutation (single-letter amino acid code) of CDK4 generates an oncogene that is resistant to normal physiologic inhibition by p16INK4a. Therefore, analysis of the structure and function of CDK4/6 may be helpful in revealing the mechanism for the increased cell generation in PV.

Other alternatives for an abnormal elevation of p16INK4amRNA in PV also exist. Li et al50 reported that p16INK4a mRNA accumulates in high levels in cells lacking Rb function and that the transcription of p16INK4a is repressed by Rb. Stott et al51 measured the expression of p16INK4a and p14ARF relative to pRb and p53 in 18 tumor cell lines and primary fibroblast strains. All 8 cell lines, which lacked Rb, had a high expression of p16INK4a. An inverse correlation of p14ARF expression and p53 status in human cell lines was also demonstrated. Thus it is possible that Rb has a mutation producing a loss of function in PV ECFCs so that less binding of E2F to Rb is present, which would result in the activation and expression of a constituency of responder genes that increase cell generation either through increased proliferation or reduced apoptosis. These genes are known to encode products necessary for S-phase progression. Because more cells would enter S phase, the molecular mechanisms controlling cell proliferation might then be activated to increase the CDK inhibitors such as p16INK4a, p27, and p14ARF. Further studies are planned along these lines which we hope will enhance our understanding of the pathogenesis of PV as well as the control of normal erythroid proliferation and differentiation.

The authors thank Maurice Bondurant for his discussions relating to this paper which were very helpful; Hong Ao and Judy Luna, for their technical assistance; and Mary Jane Rich for her expert preparation of the manuscript.

Supported by a Veterans Health Administration Merit Review Grant (to S.B.K.), National Institutes of Health (NIH) grants ROI DK-15555 and 2 T32-DK-07186 (to S.B.K.) and CA-68485 (to Vanderbilt-Ingram Cancer Center).

The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 U.S.C. section 1734.

1
Adamson
JW
Fialkow
PJ
Murphy
S
Prchal
JF
Steinman
L
Polycythemia vera: stem cell and probable clonal origin of the disease.
N Engl J Med.
295
1976
913
916
2
Green
AR
Pathogenesis of polycythemia vera.
Lancet.
347
1996
844
845
3
Heimpel
H
The present state of pathophysiology and therapeutic trials in polycythemia vera.
Int J Hematol.
64
1996
153
165
4
Hinshelwood
S
Bench
AJ
Green
AR
Pathogenesis of polycythemia vera.
Blood Rev.
11
1997
224
232
5
Michiels
JJ
Barbui
T
Finazzi
G
et al
Diagnosis and treatment of polycythemia vera and possible future study designs of the PVSG.
Leuk Lymph.
36
2000
239
258
6
Zanjani
ED
Lutton
JD
Hoffman
R
Wasserman
LR
Erythroid colony formation by polycythemia vera bone marrow in vitro.
J Clin Invest.
59
1977
841
848
7
Fisher
MJ
Prchal
JF
Prchal
JT
D'Andrea
AD
Anti-erythropoietin (EPO) receptor monoclonal antibodies distinguish EPO-dependent and EPO-independent erythroid progenitors in polycythemia vera.
Blood.
84
1994
1982
1991
8
Cotes
PM
Dore
CJ
Liu Yin
JA
Messinezy
M
Pearson
TC
Reid
C
Determination of serum immunoreactive erythropoietin in the investigation of erythrocytosis.
N Engl J Med.
315
1986
283
287
9
Dai
CH
Krantz
SB
Means
RT
Horn
WT
Gilbert
HS
Polycythemia vera blood burst-forming units-erythroid are hypersensitive to interleukin-3.
J Clin Invest.
87
1991
391
396
10
Dai
CH
Krantz
SB
Dessypris
EN
Means
RT
Jr
Horn
ST
Gilbert
HS
Polycythemia vera I: hypersensitivity of bone marrow erythroid, granulocyte-macrophage, and megakaryocyte progenitor cells to interleukin-3 and granulocyte-macrophage colony-stimulating factor.
Blood.
80
1992
891
899
11
Dai
CH
Krantz
SB
Green
WF
Gilbert
HS
Polycythemia vera III: burst-forming units-erythroid (BFU-E) response to stem cell factor and c-kit receptor expression.
Br J Haematol.
86
1994
12
21
12
Correa
PN
Eskinazi
D
Axelrad
AA
Circulating erythroid progenitors in polycythemia vera are hypersensitive to insulinlike growth factor-1 in vitro: studies in an improved serum-free medium.
Blood.
83
1994
99
112
13
Mizra
AM
Correa
PN
Axelrad
AA
Increased basal and induced tyrosine phosphorylation of the insulinlike growth factor I receptor β subunit in circulating mononuclear cells of patients with polycythemia vera.
Blood.
86
1995
877
882
14
Mizra
AM
Ezzat
S
Axelrad
AA
Insulin-like growth factor binding protein-1 is elevated in patients with polycythemia vera and stimulates erythroid burst formation in vitro.
Blood.
89
1997
1862
1869
15
Means
RT
Jr
Krantz
SB
Sawyer
ST
Gilbert
HS
Erythropoietin receptors in polycythemia vera.
J Clin Invest.
84
1989
1340
1344
16
Dai
CH
Krantz
SB
Koury
ST
Kollar
K
Polycythemia vera IV: specific binding of stem cell factor to normal and polycythemia vera highly purified erythroid progenitor cells.
Br J Haematol.
88
1994
497
505
17
Silva
M
Richard
C
Benito
A
Sanz
C
Olalla
I
Fernandez-Luna
J
Expression of Bcl-x in erythroid precursors from patients with polycythemia vera.
N Engl J Med.
338
1998
564
571
18
Dai
CH
Krantz
SB
Saywer
ST
Polycythemia vera V. Enhanced proliferation and phosphorylation due to vanadate are diminished in polycythemia vera erythroid progenitor cells: a possible defect of phosphatase activity in polycythemia vera.
Blood.
89
1997
3574
3581
19
Sui
X
Krantz
SB
Zhao
Z
Identification of increased protein tyrosine phosphatase activity in polycythemia vera erythroid progenitor cells.
Blood.
90
1997
651
657
20
Asimakopoulos
FA
Hinshelwood
S
Gilbert
JGR
et al
The gene encoding hematopoietic cell phosphatase (SHP-1) is structurally and transcriptionally intact in polycythemia vera.
Oncogene.
14
1997
1215
1222
21
Wickrema
A
Chen
F
Namin
F
et al
Defective expression of the SHP-1 phosphatase in polycythemia vera.
Exp Hematol.
27
1999
1124
1132
22
Morgan
DO
Cell cycle control in normal and neoplastic cells.
Curr Opin Genet Dev.
2
1992
33
37
23
Reddy
GP
Cell cycle: regulatory events in G1-S transition of mammalian cells.
J Cell Biochem.
54
1994
379
386
24
Vogt
PK
Reed
SI
Cyclin Dependent Kinase (CDK) Inhibitors. Current Topics in Microbiology and Immunology.
272
1998
43
148
Springer
New York, NY
25
Graña
X
Reddy
EP
Cell cycle control in mammalian cells; role of cyclins, cyclin-dependent kinases (CDKs), growth suppressor genes and cyclin-dependent kinase inhibitors (CKIs).
Oncogene.
11
1995
211
219
26
Sharpless
NE
DePinho
RA
The INK4a/ARF locus and its two gene products.
Curr Opin Genet and Dev.
9
1999
22
30
27
Serrano
M
Hannon
GJ
Beach
D
A new regulatory motif in cell-cycle control causing specific inhibition of cyclin D/CDK4.
Nature.
366
1993
704
707
28
Seranno
M
The INK4a/ARF locus in murine tumorigenesis.
Carcinogenesis.
21
2000
865
869
29
Fargnoli
MC
Chimenti
S
Keller
G
et al
CDKN2a/p16INK4 mutations and lack of p19ARF involvement in familiar melanoma kindreds.
J Invest Dermatol.
111
1998
1202
1206
30
Suzuki
H
Fujioka
Y
Nagashima
K
Cyclin D1 gene amplification and p16 gene deletion in patients with esophageal carcinosarcoma.
Diag Molec Pathol.
7
1998
253
259
31
Liew
CT
Li
H-M
Lo
K-W
et al
High frequency of p16INK4A gene alterations in hepatocellular carcinoma.
Oncogene.
18
1999
789
795
32
Belinsky
SA
Nikula
KJ
Palmisano
WA
et al
Aberrant methylation of p16INK4a is an early event in lung cancer and a potential biomarker for early diagnosis.
Proc Natl Acad Sci U S A.
95
1998
11891
11896
33
Vonlanthen
S
Heighway
J
Tschan
MP
et al
Expression of p16INK4a/p16α and p19ARF/P16β is frequently altered in non-small cell lung cancer and correlates with p53 overexpression.
Oncogene.
17
1998
2779
2785
34
Herranz
M
Urioste
M
Santos
J
et al
Analysis of the INK4a/ARF locus in non-Hodgkin's lymphomas using two new internal microsatellite markers.
Leukemia.
13
1999
808
810
35
Taniguchi
T
Chikatsu
N
Takahashi
S
et al
Expression of p16INK4A and p14ARF in hematological malignancies.
Leukemia.
13
1999
1760
1769
36
Hebert
J
Cayuela
JM
Berkeley
J
Sigaux
F
Candidate tumor-suppressor genes MTS1 (P16INK4A) and MTS2 (P15INK4B) display frequent homozygous deletions in primary cells from T- but not from B-cell lineage acute lymphoblastic leukemias.
Blood.
84
1994
4038
4044
37
Quesnel
B
Preudhomme
C
Philippe
N
et al
P16 gene homozygous deletions in acute lymphoblastic leukemia.
Blood.
85
1995
657
668
38
Quelle
DE
Zindy
F
Ashmun
RA
Sherr
CJ
Alternative reading frames of the Ink4a tumor suppressor gene encode two unrelated proteins capable of inducing cell cycle arrest.
Cell.
83
1995
993
1000
39
Sawada
K
Krantz
SB
Kans
JS
et al
Purification of human erythroid colony-forming units and demonstration of specific binding of erythropoietin.
J Clin Invest.
80
1987
357
366
40
Michiels
J
Juvonen
E
Proposal for revised diagnostic criteria of essential thrombocythemia and polycythemia vera by the thrombocythemia vera study group.
Sem Thromb Hemostas.
23
1997
339
347
41
Dai
CH
Price
JO
Brunner
T
Krantz
SB
Fas Ligand is present in human erythroid colony-forming cells and interacts with Fas induced by interferon γ to produce erythroid cell apoptosis.
Blood.
91
1998
1235
1242
42
Ross
JS
DNA ploidy and Cell Cycle Analysis in Pathology.
1
1996
Igaku-Shoin Medical Pubs
New York, NY
43
Stone
S
Jiang
P
Dayananth
P
et al
Complex structure and regulation of the p16 (MTS1) locus.
Cancer Res.
55
1995
2988
2994
44
Roussel
MF
The INK4 family of cell cycle inhibitors in cancer.
Oncogene.
18
1999
5311
5317
45
Jiang
Y
Zhao
R
Verfaillie
C
Abnormal integrin-mediated regulation of chronic myelogenous leukemia CD34+ cell proliferation: BCR/ABL up-regulates the cyclin-dependent kinase inhibitor, p27Kip, which is relocated to the cell cytoplasm and incapable of regulating cdk2 activity.
Pro Natl Acad Sci U S A.
97
2000
10538
10543
46
Weber
JD
Taylor
LB
Roussel
MF
Sherr
CJ
Bar-Sagi
D
Nucleolar Arf sequesters Mdm2 and activates p53.
Nature Cell Biol.
1
1999
20
26
47
Zhang
Y
Ziong
Y
Yarbrough
WG
ARF promotes MDM2 degradation and stabilizes p53: Arf-Ink4a locus deletion impairs both the Rb and p53 tumor suppression pathways.
Cell.
92
1998
725
734
48
Wölfel
T
Hauer
M
Schneider
J
et al
A P16INK4a – insensitive CDK4 mutant targeted by cytolytic T lymphocytes in a human melanoma.
Science.
269
1995
1281
1284
49
Zuo
L
Weger
J
Yang
Q
et al
Germline mutations in the p16INK4a binding domain of CDK4 in familial melanoma.
Nature Genetics.
12
1996
97
99
50
Li
Y
Nichols
MA
Shay
JW
Xiong
Y
Transcriptional repression of the D-type cyclin-dependent kinase inhibitor p16 by the retinoblastoma susceptibility gene product pRb.
Cancer Res.
54
1994
6078
6082
51
Stott
FJ
Bates
S
James
MC
et al
The alternative product from the human CDKN2A locus, p14ARF, participates in a regulatory feedback loop with p53 and MDM2.
EMBO.
17
1998
5001
5014

Author notes

Sanford B. Krantz, Department of Medicine-Hematology/Oncology, Vanderbilt University, 547 MRB II, 2220 Pierce Ave, Nashville, TN 37232-6305; e-mail:krantz.sanford_b@nashville.va.gov.

Sign in via your Institution