Interleukin 11 significantly increases plasma von Willebrand factor and factor VIII in wild type and von Willebrand disease mouse models

von Willebrand factor (vWf) facilitates platelet adhesion to the subendothelium, especially at high shear rates, and is the plasma carrier for factor VIII (FVIII).1-4 A quantitative or functional deficiency in vWf causes von Willebrand disease (vWd), which is the most common hereditary hemorrhagic blood disease.5-8 Symptoms of vWd include bruising, nosebleeds, bleeding gums, bleeding mucous membranes, and gastrointestinal blood loss.9 In severe vWd, joint tissue bleeds may resemble those seen in hemophilia.9 In affected females, heavy menstrual bleeding is the primary symptom.10 The gene for vWf is large (178 kilobases) and complex (52 exons), and a wide range of mutations has been described in patients with vWd.11,12 The vWf deficiencies resulting from these mutations are categorized as either qualitative or quantitative and are further classified by disease type.9,11,12 Type 1 and type 3 disease encompass mild and severe quantitative deficiency.9,11,12 Type 2 deficiency refers to qualitative abnormalities in which there is a normal level of vWf that is functionally deficient.9,11,12 The majority of patients with vWd have mild disease and are treated with desmopressin (1-deamino-8-d-arginine vasopression [DDAVP]).13,14 The benefit of DDAVP is its ability to stimulate tissue storage sites to release FVIII and vWf, resulting in increased circulating factors. Continued exposure to DDAVP results in a diminished response to the product.13,14 Patients with more severe forms of vWd and those nonresponsive to DDAVP therapy are treated with plasma concentrates containing FVIII and vWf.13 15 

Recombinant human interleukin 11 (rhIL-11, Neumega; Wyeth/Genetics Institute, Cambridge, MA) is a pleiotropic cytokine with both thrombopoietic and anti-inflammatory activity that is approved for platelet restoration following high-dose chemotherapy.16-18 During preliminary clinical trials, in addition to increases in platelet count, elevations in vWf and fibrinogen were also observed.19 Because this increase in vWf might suggest potential benefit to the vWd patient, we were interested in further evaluating this response in appropriate cell culture systems and animal models. The results presented here demonstrate that both the wild type mouse and the vWf^+/−mouse are acceptable models for studying the vWf effects of rhIL-11 treatment. Treatment with rhIL-11 produced a sustained elevation of vWf from vascular endothelial cells in both wild type and vWf^+/− mice which was similar to that observed in humans. Furthermore, the use of vWf-deficient mice provided evidence that rhIL-11 can elevate FVIII levels independently of the vWf increase, thus uncovering an additional role for rhIL-11. The results suggest that rhIL-11 has therapeutic potential in treatment of DDAVP nonresponsive mild hemophilia A and vWd patients.

We used 8-week-old C57BL/6J female mice (The Jackson Laboratory, Bar Harbor, ME). The vWf heterozygous and vWf-deficient mice were on a mixed 129SV/C57BL/6J background as described previously20and were used between 6 and 10 weeks of age. Four mice were housed in each cage and fed a standard chow diet and water ad libitum. All animal experiments were approved by the Animal Care and Use Committee of the Center for Blood Research, Boston, MA.

Placebo (rhIL-11 vehicle buffer) and endotoxin-free rhIL-11 were prepared at Wyeth/Genetics Institute, Cambridge, MA.21Mice were subcutaneously injected with placebo or rhIL-11 diluted in phosphate-buffered saline (PBS) (250 μg/kg, unless specified otherwise) once a day for 4, 7, 14, or 21 consecutive days.22 In another set of experiments mice received an intravenous (IV) injection into the tail vein with placebo or a bolus dose of 250 or 1000 μg/kg rhIL-11.

Mice were anesthetized with Metofane (Schering-Plough Animal Health, Union, NJ), and blood was collected by retro-orbital venous plexus sampling in polypropylene eppendorf tubes containing ethyelendiamine tetraacetic acid (EDTA). Complete blood counts were determined using an automatic cell counter (Coulter Electronics, Hialeah, FL).

Mice were anesthetized, and blood was obtained by retro-orbital venous plexus sampling in polypropylene eppendorf tubes containing 129 mM trisodium citrate. Plasma was prepared by centrifugation of the blood at 2500g for 15 minutes at room temperature. We measured vWf antigen level using an enzyme-linked immunosorbent assay (ELISA) technique as described.20 The number of units in each sample was determined based on the absorbance value of a normal pooled plasma obtained from 10 wild type mice. We defined normal pooled plasma as containing one unit vWf antigen per mL plasma. For the in vitro studies with human umbilical vein endothelial cells (HUVECs), vWf antigen level was determined in the culture media, and human plasma was used as the reference plasma (Sigma Chemical, St Louis, MO).

FVIII concentration was determined using a one-stage clotting assay. Plasma was harvested as described for the vWf assay and kept frozen until the assay. Plasma was diluted 1:40 with Owren's buffer (Sigma), and normal reference plasma (Sigma) was used as a standard. FVIII-deficient plasma (Helena Laboratories, Beaumont, TX) and diluted mouse plasma were mixed in a glass tube with activated partial thromboplastin time reagent (Instrumentation Laboratory, Lexington, MA). The tube was then transferred to a 37°C water bath, and after 5 minutes, clotting was initiated by the addition of 25 mM calcium chloride. Clotting was monitored visually during constant mixing in the 37°C water bath, and the time necessary for the clotting to occur was recorded. FVIII activity in the samples was determined using the standard curve obtained with the reference plasma.

Heart, lungs, liver, and spleen were collected after 1 or 4 days from wild type mice injected subcutaneously with 250 μg/kg/d rhIL-11. Total RNA was extracted from liquid nitrogen snap-frozen tissues using the Tri-Reagent (Molecular Research Center, Cincinnati, OH) according to manufacturer's specification. In brief, 100 mg tissue was homogenized in 1 mL Tri-Reagent and stored at room temperature for 5 minutes. One-tenth volume of 1-bromo-3-chloropropane (BCP) was added, shaken vigorously for 15 seconds, and kept at room temperature for 10 minutes prior to centrifugation at 12 000g for 15 minutes at 4°C. The aqueous phase was removed and placed into a new tube, where RNA was precipitated for 10 minutes at −20°C using 0.5 mL isopropanol. RNA was collected by centrifugation at 12 000gfor 10 minutes at 4°C. The supernatant was removed, and RNA was washed with 1 mL 75% ethanol and then followed by subsequent centrifugation at 12 000g for 10 minutes. Ethanol was removed, and RNA was allowed to dry for 5 minutes. RNA was resuspended in diethylpyrocarbonate-treated (DEPC) sterile water. Total RNA was treated with RQ1 DNase I (Promega, Madison, WI) and RNase inhibitor (Five Prime Three Prime, Boulder, CO) for one hour at 37°C. RNA clean was performed using Qiagen Rneasy Minicolumns (Qiagen, Valencia, CA) according to manufacturer's specification. Recombinant thermostable DNA polymerase was used to reverse transcribe and amplify 125 ng total RNA using the Perkin Elmer Taqman EZ RT-PCR (reverse transcriptase–polymerase chain reaction) kit (Perkin Elmer Applied Biosystems, Foster City, CA) with gene-specific forward and reverse primers and a fluorescently labeled probe at the 5′ end with 6-carboxy-flourescein (6-FAM).23-25 Primer and probe sequences were generated using Primer Express software (Perkin Elmer).

Murine FVIII forward primer 5′-TGTTTGTAGAGTACAAGGACCAGCTT-3′, reverse primer 5′-GAACCTCAGTCCAAATGGTAGGA-3′, and Taqman probe 5′-TGCCAAGCCCAGGCCACCCT-3′ were used. Murine vWf forward primer 5′-ATCCGCGTGGCAGTGGTA-3′, reverse primer 5′-CGCTTCCGGGCCTTG-3′, and Taqman probe 5′-CATGATGGCTCCCGTGCCTACCTTG-3′ were used. Duplicate samples were reverse transcribed for 30 minutes at 60°C, followed by 40 cycles of amplification for 15 seconds at 95°C and one minute at 60°C using the ABI Prism 7700 sequence detection system (Perkin Elmer) as described by the manufacturer. Gene-specific amplification was detected as a fluorescent signal during the amplification cycle. Gene-specific message quantification was evaluated by fluorescence intensity levels of unknown samples compared to fluorescence intensity of known messenger RNA (mRNA) levels. Amplification of a housekeeping gene, murine glyceraldehyde 3 phosphate dehydrogenase (GAPDH), was performed on all samples to account for RNA-level variations. All genes were normalized to GAPDH mRNA levels, and levels of gene-specific messages were depicted as normalized Taqman units as determined by standard curve.

Blood was collected as described in the vWf ELISA, except that blood samples were placed in polypropylene tubes containing 0.1 volume of 38 mM citric acid, 75 mM trisodium citrate, and 100 mM dextrose. The P-selectin ELISA was completed as described by Hartwell et al.26 

HUVECs (second or third passages) were grown as described.27 To visually assess the release of storage granules by rhIL-11, cells were grown on glass coverslips and incubated for 30 minutes, 3 hours, or 24 hours at 37°C with 1, 10, 50, and 1000 ng/mL rhIL-11 added to the media. Control cells were treated with rhIL-11 diluting solution (PBS). For vWf ELISA, the HUVECs were grown in 35-mm dishes (Falcon), and after confluency, cells were rinsed twice with fresh media and treated with 100 or 1000 ng/mL rhIL-11. After the designated incubation time, media were collected, centrifuged, and stored at −20°C prior to vWf assay.

Cells were grown on coverslips, fixed in 3.7% (vol/vol) formaldehyde, and permeabilized with 0.5% Triton X-100 in PBS. The cells were stained for vWf using rabbit polyclonal antibody to vWf (American Bioproducts, Parsippany, NJ) at 1:100 dilution followed by fluorescein-conjugated sheep antibody to rabbit immunoglobulin G (IgG) (ICN Pharmaceuticals, Costa Mesa, CA) at 1:1000 dilution. All incubations with antibodies were performed for 30 minutes at 37°C.

Plasma samples diluted 1:20 in 10 mM Tris-HCl (tris[hydroxymethyl aminomethane–hydrochloride) and 1mM EDTA (pH 8) containing 2% sodium dodecyl sulfate (SDS) were layered on 1% agarose gels (agarose IEF, Amersham Pharmacia Biotech, Piscataway, NJ) poured on Gelbond (FMC, Rockland, ME). Gel electrophoresis was run horizontally at a constant amperage of 10 mA. The gel was fixed in 25% isopropyl alcohol and 10% acetic acid for 30 minutes and rinsed. Multimers were revealed by incubating the dried Gelbond with a polyclonal antibody to human vWf labeled with iodine 125 (¹²⁵I) and visualized by autoradiography.

We injected 8-week-old female vWf^−/− mice subcutaneously with a placebo or 250 μg/kg/d rhIL-11 for 7 days. The mice were anesthetized, and hair from the internal side of the hind leg was removed with electric clippers. A standard incision (5-mm long by 1-mm deep) was made with an automatic device (Surgicutt; ITC, Edison, NJ). Every 15 seconds a filter paper was applied to the wound for 5 seconds until the blood no longer stained the paper. Bleeding time was determined to the nearest 15 seconds.

vWf^−/− mice were injected subcutaneously with a placebo or 250 μg/kg/d rhIL-11 for 5 days. On the fifth day following rhIL-11 injection, the mice were injected intravenously with 40 U/kg rhvWf (Wyeth/Genetics Institute). Blood was collected from 4 mice injected with rhIL-11 and from 4 mice injected with placebo immediately after the rhvWf injection and at various time points following the injection: 30 minutes and after 3, 6, 14, and 24 hours. Each mouse was bled only once except for the 14- and 24-hour time point, where the same set of mice was bled twice. Plasma vWf antigen levels were determined by ELISA.

Data are presented as the mean ± SEM. Statistical significance was calculated using Student t test for all analyses. Unpaired or paired tests were performed as indicated in the figure legends.

C57BL6J mice were injected subcutaneously with a placebo or 250 μg/kg/d rhIL-11 for a total of 7 days. A group of mice was bled and killed at day 4 and another on day 7. The remaining animals were killed on day 14, 7 days after the last injection. As an internal control for rhIL-11 activity, platelet counts were measured in the peripheral blood in all mice. Similar to the preliminary report on humans,19 the number of platelets was significantly increased in the rhIL-11–treated mice at day 4 and day 7 (Figure1A). Platelet numbers declined after treatment was stopped and returned to preinjection levels by day 14. The white blood cell count and hemoglobin level did not change in any group during the evaluation period. Plasma vWf was increased 2.0- and 1.8-fold at day 4 and 7, respectively, in the rhIL-11–treated mice compared to the vehicle-injected mice (P = .005 andP = .01, respectively) (Figure 1B). Similar to the platelet counts, the vWf levels returned to pre-injection levels by day 14. FVIII levels were increased the most by rhIL-11 administration (Figure 1C). The rhIL-11 injection enhanced FVIII level 2.6-fold at day 4 and 2.9-fold at day 7 compared to day 0. These increases were highly significant (P = .0002) when compared to preinjection levels or to placebo-injected mice (P < .0001). Interestingly, FVIII level at day 14 was still significantly higher in the rhIL-11–treated mice than in the controls (P < .04). We also tested whether the effect of rhIL-11 could be long-lasting. For this purpose we injected the mice with rhIL-11 daily for up to 21 days, and we found that the elevation in platelet numbers and vWf and FVIII levels was sustained for as long as the treatment lasted (not shown).

The multimeric composition of vWf from wild type mice injected with rhIL-11 for 7 days was similar to that of mice injected with placebo (Figure 2). We could not observe an increase of the higher molecular weight multimers characteristic of vWf stored in the endothelial cell storage granules (Weibel-Palade bodies)25 after rhIL-11 treatment.

To determine if lower doses of rhIL-11 were active, we compared the effect of 3 doses of rhIL-11 (30, 100, and 300 μg/kg) in wild type mice. The 2 highest concentrations, 100 and 300 μg/kg, produced significant elevations in platelet numbers and in vWf and FVIII levels (Figure 3A-C), with the 300 μg/kg dose being the most effective. At 30 μg/kg, rhIL-11 did not reliably induce an increase in platelet numbers and vWf and FVIII levels. Thus, a dose of 100 μg/kg is necessary to achieve biological activity in mice.

To investigate whether rhIL-11 treatment directly induces release of Weibel-Palade bodies containing vWf in endothelial cells, we tested the effect of rhIL-11 on vWf release by HUVECs in vitro. HUVECs were grown to confluency and treated with placebo or rhIL-11 at concentrations of 0-1000 ng/mL for different incubation times (30 minutes to 72 hours). We evaluated the release of Weibel-Palade bodies in 2 different ways. First, we stained treated HUVECs by immunofluorescence with an antibody to vWf to visualize Weibel-Palade bodies. The rhIL-11 treatment of HUVECs did not result in a loss of Weibel-Palade bodies, and we could not detect the typical release patches28 of cell-surface–associated vWf (data not shown). Second, we collected HUVEC culture medium after treatment with or without rhIL-11 and measured vWf levels by ELISA (Table1). Again, we did not observe any rhIL-11–dependent vWf secretion.

To determine whether rhIL-11 could stimulate Weibel-Palade–body secretion in vivo, we injected 250 or 1000 μg/kg rhIL-11 via the tail vein in wild type mice and measured vWf and FVIII levels 3 hours after injection. Three hours was estimated to be sufficient time for the secretion of the granules to occur when endothelial cells were treated by secretagogues, but it was not long enough for significant de novo synthesis of the vWf protein.27 Platelet counts and vWf and FVIII concentrations in the plasma were not changed by rhIL-11 treatment (Table 2). White blood cell count and hemoglobin level were also unchanged. Thus rhIL-11 does not appear to induce vWf or FVIII release from storage compartments.

P-selectin, like vWf, is located both in the Weibel-Palade bodies of endothelial cells and the α granules of platelets, and therefore P-selectin is expressed on the surface of these cells by the same secretagogues or agonists that stimulate vWf release. Soluble P-selectin in the plasma is generated by shedding of the membrane form of P-selectin and usually reflects platelet activation.29We tested the effect of daily subcutaneous injections of 250 μg/kg rhIL-11 on the level of soluble P-selectin in wild type mice. The conditions chosen were the ones found to be optimal for vWf increase. There was no significant increase in P-selectin levels observed after rhIL-11 (Table 3). This result indicates that excessive platelet activation by rhIL-11 did not contribute significantly to the vWf increase.

To investigate whether an rhIL-11 effect on vWf and FVIII levels occurred through a stimulation of their synthesis, we used a quantitative method to determine the amount of vWf and FVIII mRNA after injection of wild type C57BL/6J mice with a placebo or 250 μg/mL rhIL-11 for 4 days. We confirmed the significant increase in platelet counts and in vWf and FVIII levels in rhIL-11–injected mice and measured mRNA for vWf in lung, kidney, and spleen and for FVIII in liver, kidney, and spleen. There were no differences detected in vWf and FVIII message between rhIL-11–treated mice and controls (Figure4). The same result was obtained when we looked at an earlier time point (24 hours after the beginning of the rhIL-11 injections) (data not shown). Thus, rhIL-11 does not appear to lead to increased synthesis or stabilization of vWf and FVIII mRNA.

To determine whether rhIL-11 would be able to produce an increase of vWf and FVIII in a model of vWd type 1, we used vWf-heterozygous mice, which have only 50% vWf antigen level and approximately 60% FVIII compared to wild type mice.20 The vWf-heterozygous mice injected for 7 days with 250 μg/kg rhIL-11 responded in a similar fashion to wild type mice (Figure5). Platelet numbers and vWf and FVIII levels were significantly increased in the rhIL-11–treated group compared to the placebo-treated group. A slight increase was seen in the placebo-treated mice, likely due to the stress of daily injections. After rhIL-11 treatment of vWf^+/− mice, vWf levels increased very close to the baseline level observed in wild type mice (Figure 5). Strikingly, FVIII level after rhIL-11 treatment of the heterozygous mice was higher than the baseline level seen in wild type mice (P = .0004). Using these vWf^+/− mice, we also performed a long-term rhIL-11 injection study and observed that the high levels of vWf and FVIII could be sustained at their maximum for as long as the treatment was continued, ie, 21 days of daily subcutaneous injections of 100 μg/kg rhIL-11 (data not shown).

In the vWf-deficient mice, the level of FVIII is about 20% that of wild type mice.20 We then questioned whether rhIL-11, in the complete absence of vWf, would still be able to increase FVIII levels. After injecting the vWf^−/− mice with 250 μg/kg rhIL-11 subcutaneously for 7 consecutive days, platelet numbers increased in a fashion similar to those in wild type mice. Surprisingly, as in vWf^+/+ and vWf^+/− mice, we noticed a significant increase in FVIII levels, up to the level of untreated vWf^+/− mice, after rhIL-11 treatment of vWf^−/− mice (Figure 6). Therefore, it appears that the FVIII increase observed after rhIL-11 treatment is not simply a consequence of the increase in vWf. In terms of absolute values, the FVIII level in vWf^−/− mice after rhIL-11 treatment was lower than the FVIII level in vWf^+/+and vWf^+/− mice. This suggests that without the concomitant increase in vWf, FVIII level cannot be optimally elevated, probably due to an increased rate of clearance. However, the FVIII level was increased to 0.8 U/mL, which is similar to the baseline level observed in vWf^+/− mice. To test whether rhIL-11 injections could improve the hemostatic status of vWf^−/−mice, we measured bleeding time in these mice using a skin model. We found that rhIL-11 injections were able to significantly reduce the skin bleeding time of vWf^−/− mice compared to placebo injections (115 ± 12 seconds vs 265 ± 37 seconds, respectively) (Figure 7).

To investigate whether rhIL-11 induces increase in vWf levels through a prolongation of its half-life, we measured clearance of rhvWf injected in vWf^−/− mice treated with placebo or rhIL-11. We did not detect any differences in the clearance rate of rhvWf between and placebo- and rhIL-11–treated mice, with a half-life of about 2 hours in each group (Figure8).

Thrombocytopenia is associated with a number of pathological situations or can arise as a complication of medical treatments such as chemotherapy. As a consequence, numerous studies have focused on isolating growth factors that are able to stimulate megakaryocyte development and platelet production. Several of these factors have been identified including IL-6,30IL-11,32 and thrombopoietin.31 During the preliminary trials evaluating the potential of IL-11 in reducing chemotherapy-induced thrombocytopenia, it was observed that rhIL-11 increased vWf and fibrinogen levels.19 Because of the potential therapeutic interest of this effect in preventing bleeding complications, we decided to confirm this observation in mice and investigate the mechanisms involved. We found that rhIL-11 treatment of wild type mice for 7 days resulted not only in elevation of vWf levels but also in FVIII levels. The vWf increase was rapidly reversible on discontinuation of rhIL-11 treatment. However, we noted that the FVIII level remained elevated even after the vWf level had returned to baseline. Similar observations about the kinetics of FVIII increase have been reported following infusions of plasma or cryoprecipitates in vWd patients.33 34 

IL-11 is a member of the IL-6 family of cytokines that includes ciliary neurotrophic factor, leukemia inhibitory factor, oncostatin M, and cardiotrophin.35-37 While each of these cytokines uses the gp130 chain as part of their receptor complex, the high-affinity binding is determined by the association of this common chain with a cytokine-specific α chain.37,38 The IL-11 receptor α chain is widely distributed in the mouse and human in the heart, lung, liver, kidney, spleen, and bone marrow.38 

The effect of IL-11 on the expression of acute phase proteins was first reported in the liver. Indeed, exposure of hepatocytes to rhIL-11 induces expression of fibrinogen, α1 antitrypsin, and α2 macroglobulin.39 Furthermore, an elevation of acute-phase proteins has also been reported following systemic administration of rhIL-11 to experimental animals, human volunteers, and patients.16-19,22 Our present results confirm and extend these findings. vWf can be considered as an acute-phase protein because its level is elevated with exercise, during pregnancy, and under certain hormonal stimulation.40 However, in contrast to the direct effect of IL-11 observed on rat hepatocytes, treatment of HUVECs with rhIL-11 did not result in the release or elevation of vWf. A recent study identified the IL-11 receptor α chain in cultured endothelial cells and showed that endothelial cells can respond to rhIL-11 treatment by induction of signaling pathways.41Therefore, the absence of a direct release of vWf from HUVECs could suggest that either the HUVECs do not have the regulatory elements necessary for IL-11 action or, more likely, that an indirect mechanism is acting to promote vWf secretion. Unlike DDAVP in humans, rhIL-11 does not induce a rapid elevation of vWf or FVIII following a single high-dose administration, which suggests that the IL-11–induced vWf increase is not the result of the release of Weibel-Palade bodies. Also the plasma vWf multimeric pattern observed after rhIL-11 treatment did not show the high molecular weight multimers characteristic of stored vWf.28 

The origin of the increase in plasma vWf was next investigated. Plasma vWf usually comes almost exclusively from the endothelial cells,42 but we could not exclude the possibility that the stimulated bone marrow megakaryocytes and/or the increased circulating platelets could also be a source of vWf. However, we could not detect any increase of soluble P-selectin after rhIL-11 treatment, which would indicate α-granule release following platelet activation.29 The absence of rhIL-11 effect on platelet activation was reported previously in humans17 as well as the fact that IL-11 does not increase vWf protein content in the α granules of megakaryocytes or platelets in mice.43 We also confirmed this result in our experiments (not shown). Taken together, these results suggest that platelets are an unlikely source of the increased vWf observed in this study.

Taking advantage of our mouse model of vWd, we tested the effect of rhIL-11 in this animal model. Studies performed with the vWf heterozygous mice provided evidence that rhIL-11 can bring vWf and FVIII levels up to normal levels in a model of type 1 vWd. Similar to the wild type mice, the elevation of FVIII in the vWf heterozygous mice could be explained by the availability of more carrier sites provided by the elevated vWf levels. Our results obtained with the vWf-deficient mice were surprising because in these animals, rhIL-11 treatment resulted in a significant elevation in plasma FVIII in the absence of vWf, which indicates that to some extent, FVIII can be regulated independently of vWf.

We then investigated whether vWf or FVIII mRNA was increased after rhIL-11, which would indicate either a stimulation of the synthesis or an increased stability of the mRNA. Under the same conditions that resulted in protein level increase, we could not detect a change in mRNA levels. In an attempt to identify the mechanism(s) underlying rhIL-11–mediated increase of vWf and FVIII protein expression, we measured the clearance rate of rhvWf injected in vWf-deficient mice after rhIL-11 treatment. Proteolytic degradation is likely to contribute to vWf clearance, and a vWf degrading protease enzyme was partially purified and characterized from human plasma.44 The rhIL-11–mediated vWf increase could therefore possibly be mediated through a reduced activity of this protease. However, we did not detect any prolongation of vWf half-life following rhIL-11 treatment. This result does not preclude a possible effect of rhIL-11 on FVIII half-life. In the vWf-deficient mice, an in vivo prolongation of FVIII half-life can be achieved by the receptor-associated protein (RAP), which inhibits the clearance mediated by the low-density lipoprotein receptor-related protein.45 A stimulatory effect of rhIL-11 on the RAP protein is possible.

The most plausible hypothesis based on our results is that rhIL-11 plays a role at the translational level. Indeed there is growing evidence demonstrating the importance of RNA-protein interactions in the regulation of mRNA translation, which results in different amounts of protein produced.46,47 There are a few examples of interleukins acting at the translational and/or post-translational level. IL-6 can inhibit the expression of the syndecan-1 proteoglycan without any effect on the mRNA levels.48 Similarly, IL-10 can modulate intercellular adhesion molecule (ICAM)-1 expression by glial cells by inhibiting the increase in protein expression mediated by a number of other cytokines, but it has no effect on the accompanying increase in mRNA.49 

Our results demonstrate that rhIL-11, in addition to its effect on platelet production, can significantly increase FVIII levels in a mouse model of severe vWd and can improve hemostasis in these mice. A significant increase of vWf and FVIII levels can also be obtained in wild type mice and in heterozygote mice representing a model of type 1 vWd. Importantly, this increase can be sustained for long periods of time, in contrast to the effect observed with DDAVP, where patients treated repeatedly may become less responsive, perhaps due to the depletion of the storage granules.50 These results suggest that rhIL-11 might be helpful in alleviating the bleeding symptoms in vWd. Treatment with rhIL-11 might prove to be especially useful in those clinical situations in which vWf and/or FVIII levels must be maintained above baseline for a prolonged period of time. For example, it is possible to imagine a preventive use of rhIL-11 treatment before surgical procedures in order to achieve an optimal hemostatic balance in patients suffering from mild hemophilia A or vWd.

We wish to thank Charlene DeClercq for technical help, Sangeetha Subbarao and Jessie Papalia for mouse husbandry, and Lesley Cowan for assistance with preparation of the manuscript. We also would like to thank Drs Richard Hynes and Bruce Ewenstein for reading the manuscript and for helpful discussions.