Molecular Targets for Therapy (MTT)

Hyperforin inhibits MMP-9 secretion by B-CLL cells and microtubule formation by endothelial cells


We previously reported that hyperforin (HF), a natural phloroglucinol purified from Saint John's wort, can induce the apoptosis of leukemic cells from patients with B-cell lymphocytic leukemia (B-CLL) ex vivo. We show here that treatment of cultured B-CLL patients' cells with HF results in a marked inhibition of their capacity to secrete matrix metalloproteinase-9, an essential component in neo-angiogenesis through degradation of the extracellular matrix process. The phloroglucinol acts by decreasing the production of the latent 92 kDa pro-enzyme. The inhibitory effect of HF is associated with a decrease in VEGF release by the leukemic cells. Moreover, HF is found to prevent the formation of microtubules by human bone marrow endothelial cells cultured on Matrigel, evidencing its capacity to inhibit vessel formation. Our results show the antiangiogenesis activity of HF and strengthen its potential interest in the therapy of B-CLL.


The role of angiogenesis in the growth and survival of leukemic cells has been recently demonstrated by showing that the progression of several forms of leukemia is related to the degree of angiogenesis.1 By measuring microvessel and hotspot densities in bone marrow trephine biopsies, Kini et al.2 showed that both parameters are higher in patients with B-cell lymphocytic leukemia (B-CLL) than in normal individuals, indicating that angiogenesis may be involved in the pathogenesis of this leukemia. Similar results were obtained by Wolowiec et al.3 who also found that B-CLL bone marrow displays a higher proliferative activity than normal bone marrow, as assessed by the mean number of nucleolar organizer regions. These data indicate that neo-angiogenesis occurs in B-CLL, despite being a hematologic disorder.

Matrix metalloproteinase-9 (MMP-9) is an important component of neo-angiogenesis, notably by its capacity to degrade some components of the extracellular matrix, a step required for the sprouting of new blood vessels (reviewed in Van den Steen et al.4). We reported previously that B-CLL leukemic cells express and release the pro-form of MMP-9,5 Molica et al.6 made similar constatations and found that the serum level of MMP-9 in B-CLL patients is predictive of their outcome. High levels of intracellular MMP-9 were related to disease severity and poor patient survival, and MMP-9 associated with areas of tissue invasion and remodelling. In addition, MMP inhibitors reduce the migration of B-CLL cells through endothelial monolayers, indicating that MMP-9 could play a role in malignant cell entry and egress to and from involved tissue.7 Together, these data raised the possibility that a strategy based on MMP-9 modulation may have a therapeutic potential for advanced B-CLL. Owing to this purpose, the identification of agents capable of inhibiting the production of MMP-9 appears of interest.

Extracts from the herb Hypericum perforatum (St John's wort) have been used in traditional medicine for centuries, notably for the treatment of depression. Several biologically active compounds have been isolated and characterized from these extracts, including naphthodianthrones such as hypericin, flavonoids such as rutin, quercetin, quercitrin and biapigenin, and phloroglucinols such as hyperforin (HF).8 Biochemical studies suggested that HF is one of the main active constituents of St John's wort and is a key mediator of the antidepressant effect inasmuch as it inhibits the synaptosomal uptake of 5-HT by elevating free intracellular [Na+]i concentration9 through the activation of non-selective cation channels.10 Furthermore, HF displays antibacterial11 and antioxidant12 properties, and is an inhibitor of 5-lipooxygenase, cyclooxygenase-113 and various cytochrome P450 isoforms.14 As regards the immune system, HF inhibits the proliferative response of PBMC in the absence of toxic effects; in addition, exposure of endothelial cells to HF results in suppression of alloreactive T-cell proliferation.15 Moreover, HF was shown to inhibit cell growth by inducing apoptosis in myeloid leukemia cell lines,16 and similar results were observed for human tumors in vivo.17 Recently, we demonstrated that HF is a stimulator of apoptosis in cultured B-CLL patients' cells.18

These data prompted us to investigate the in vitro effects of HF on several parameters involved in B-CLL angiogenesis. The natural phloroglucinol was found to inhibit the production of MMP-9 and VEGF by B-CLL patients' cells and to disrupt the formation of pseudo-tubules by endothelial cells derived from human bone marrow. Our results show the antiangiogenic properties of HF and suggest new strategies for the therapy of B-CLL.

Materials and methods

Patients, cells and cell cultures

Peripheral blood samples from patients with B-CLL were provided by the Hematology Department of Hôtel-Dieu hospital (Paris). The samples were collected after informed consent of the patients, in agreement with the rules of the revised Helsinki protocol. Diagnosis was established according to the standard clinical and international workshop on CLL (IWCLL) criteria, including lymphocyte morphology and co-expression of CD5, CD20 and CD23 antigens. A total of 23 patients were selected, including 8 women and 15 men with a mean age of 67.3±10.9 years (range: 34–85 years). The time since diagnosis varied between 0 and 10 years. Patients were randomly chosen for each type of experiment because CD38 and ZAP-70 expression, cytogenetics and mutational VH status were available for only a fraction of them, thus hampering a risk-group analysis. Leukemic B-cells were purified from B-CLL blood samples as described previously.19 Freshly isolated B-CLL cells were cultured at a seeding density of 2 × 106 cells/ml in RPMI-1640 medium supplemented with 2 mM fresh glutamine, 1 mM sodium pyruvate, 100 IU/ml penicillin, 100 μg/ml streptomycin and 10% FCS (PAA Laboratories, Les Mureaux, France) at 37°C in an incubator containing 5% CO2.

The human bone marrow endothelial cell (HBMEC) line was established and provided by Dr K Pienta.20 These cells were routinely maintained in culture in MCDB131 medium (Gibco BRL, Life Technologies, Cergy Pontoise, France) supplemented with 2 mM glutamine, 1 mM sodium pyruvate, 100 IU/ml penicillin, 100 μg/ml streptomycin and 10% FCS (Myoclone super plus, Gibco BRL) and with the attachment factor from Cascade biologics (Mansfield, UK) at 37°C in an incubator containing 5% CO2. Subcultures were performed using trypsin for cell detaching.


Hyperforin was purified according to the methods described elsewhere.21 Two different batches were used and stock solutions (1 mg/ml) were prepared in ethanol. The broadly specific caspase inhibitor z-VAD-fmk was purchased from Biomol (Plymouth Meeting, PA, USA). The other reagents and chemicals were from Sigma (St Louis, MO, USA).

Matrix metalloproteinase-9, VEGF and bFGF production assays

B-cell lymphocytic leukemia cells seeded at 2 × 106/ml were cultured for 48 h and cell-free culture supernatants were harvested. The production of MMP-9 by B-CLL cells was assayed by measuring the amounts of MMP-9 released in the supernatants using two distinct ELISA. One detects the total MMP-9 concentration (both the latent pro-enzyme and the mature active MMP-9) and is based on absorbance measurements (Quantikine, R&D Systems, Abingdon, UK). The other ELISA is specific for the active mature form of MMP-9 and is based on fluorescence detection (Fluorokine, R&D Systems). VEGF and bFGF concentrations were also assayed using ELISA kits according to the manufacturer's recommendations (R&D Systems).

Zymographic analysis of MMP gelatinase activity

MMP activity was studied by zymography, which reveals the gelatinase activity of latent pro-enzymes (zymogenes) in addition to that of mature MMPs, according to a method described previously.5 Briefly, equal amounts of the supernatants (20 μl) mixed with Laemmli sample buffer lacking β-mercaptoethanol were electrophoresed on 7.5% SDS-PAGE containing 0.1% gelatin. Gels were washed and incubated overnight at 37°C, and further stained with Coomassie Blue, then destained. Gelatinase activities of MMPs were detected as clear bands on the background of Coomassie Blue-stained gelatin. The NIH Image 1.44b11 software was used for the analysis of band intensities after acquisition with a digital imager (Vilber Lourmat, Marne-la-Vallée, France). The recombinant pro-MMP-9 (Oncogene, Calbiochem, La Jolla, USA) was used as a positive control.

ELISpot assay

The human MMP-9 enzyme-linked immunospot (ELISpot) assay (R&D Systems) is designed to detect MMP-9-secreting cells at the single-cell level. Based on a quantitative sandwich ELISA, it detects both the pro- and mature forms of MMP-9. Briefly, 5 × 104 cells were distributed in wells of PVDF-backed microplates pre-coated with an anti-human MMP-9 mAb for 12 h of culture. During this incubation, the immobilized antibody in the immediate vicinity of the cells binds the released MMP-9. After washings, a biotinylated polyclonal antibody specific for human MMP-9 was added. The wells were washed and incubated with streptavidin-conjugated alkalinephosphatase, and finally washed again before adding a substrate solution (BCIP/NBT). Blue–black colored precipitates formed at the sites of protease localization appeared as spots in light microscopy, with each individual spot corresponding to a single cell secreting MMP-9.

Western blot experiments

Matrix metalloproteinase-9 expression was analyzed at the protein level by Western blotting, according to a modification of the technique already detailed.22 Briefly, cells were lysed with RIPA buffer for 30 min at 4°C. Equal amounts of the cell lysates (50 μg of proteins) were immunoprecipitated for 18 h with 2 μg sheep polyclonal anti-MMP-9 antibody (Biogenesis, Pool England, UK) and a mix of A and G Sepharose beads (Santa-Cruz Biotechnologies, CA, USA) according to the method of Kamiguti et al.7 The immunoprecipitated proteins were separated on 10% SDS-PAGE in non-reducing conditions and then electrotransferred overnight at 4°C on Immobilon-P membranes (Millipore, Bedford, MA, USA). The membranes were saturated with 5% non-fat milk in TN buffer (10 mM Tris, pH=8, 200 mM NaCl) containing 0.1% Tween 20, and further incubated with a mouse anti-MMP-9 mAb (Ab-1, Oncogene, Calbiochem) in TN buffer containing 3% non-fat milk and 0.05% Tween 20. The immunoblotted proteins were revealed with a HRP-goat anti-mouse Ig by chemiluminescence (ECL, Amersham Pharmacia, Uppsala, Sweden).

RT-PCR analysis of mRNA expression

RNA extraction from B-CLL cells and subsequent cDNA synthesis were performed as described previously.23 The cDNAs for human MMP-9 and β2-microglobulin (used to normalize the PCR products) were amplified by PCR as follows: MMP-9 cDNA (296 bp) was amplified using the sense primer 5′-IndexTermGGA GAC CTG AGA ACC AAT CTC-3′ and the antisense primer 5′-IndexTermTCC AAT AGG TGA TGT TGT CGT-3′; β2-microglobulin cDNA (165 bp) was amplified using the sense primer 5′-IndexTermCAT CCA GCG TAC TCC AAA GA-3′ and antisense 5′-IndexTermGAC AAG TCT GAA TGC TCC AC-3′. The PCR products were visualized by electrophoresis in 2% agarose gel containing 0.2 μg/ml ethidium bromide, and analyzed with the NIH Image 1.44b11 software.

Tubule formation assay

The formation of microtubules was assayed using methods described previously.24 Briefly, HBMECs were detached with accutase (PAA Laboratories), which protects the adhesion molecules, then washed and counted. Cells were further resuspended in fresh medium and layered into microplate wells (3 × 105 cells/well) previously coated with 36 μl Matrigel-Rich™ (BDbiosciences, Grenoble, France) for 45 min at 37°C. The wells were photographed at different times, using a standard light microscope. Tubular structures were analyzed by Microvision Saisam software.24

DNA fragmentation assay

DNA fragmentation, a nuclear event of apoptosis, was quantified by detection of cytoplasmic histone-associated DNA fragments (mono- and oligonucleosomes) from cytosols using an ELISA with anti-histone and anti-DNA fragment mAbs (Cell Death Detection ELISAPLUS, Roche Diagnostics, Indianapolis, IN, USA), as described previously.22

Statistical analysis

Cells were cultured in duplicates or triplicates and each experiment repeated 2–5 times for each cell preparation. Statistical analyses were performed using the Statview software. The unpaired two-tailed t-test, as modified for small samples, was used for the comparison of test and control groups.


Hyperforin inhibits the production of matrix metalloproteinase-9 by B-cell lymphocytic leukemia patients' cells

Using an ELISA measuring total MMP-9 (latent inactive pro-form plus mature active enzyme), we found that leukemic cells from most B-CLL patients secreted elevated concentrations of MMP-9 in their culture supernatants: a mean value around 10 ng/ml (N=19) was observed after 48 h of culture by seeding 2 × 106 cells/ml (Table 1a). In contrast, when an ELISA specific for the active mature form of MMP-9 was used, only very low levels were detected (mean 0.40 ng/ml±0.21; N=4), as shown in Table 1b. These data indicate clearly that B-CLL cells released almost exclusively the latent form of MMP-9 (pro-MMP-9), in agreement with previous reports, including ours.5, 6 This was confirmed by pre-incubating the supernatants with aminophenylmercuric acid (APMA), an activator of pro-MMP-9, which increased markedly the concentrations of active MMP-9, thus revealing the presence of high levels of the latent pro-MMP-9 in the supernatants (Table 1b). Upon treatment with HF (5 μg/ml), the levels of total MMP-9 were significantly reduced, while those of the mature active enzyme remained unchanged as compared to untreated controls (Table 1a and b). In addition, the high levels of mature active MMP-9 in APMA-activated supernatants were also significantly decreased after exposure to HF (Table 1b). These results therefore indicate that HF treatment of cultured B-CLL cells inhibits their secretion of the latent pro-MMP-9.

Table 1 Hyperforin inhibits MMP-9 production by B-CLL patients' cells

Effects of hyperforin on gelatinase activity of secreted matrix metalloproteinase-9

Gelatinase activity of MMP-9 produced by B-CLL cells was analyzed by zymography of culture supernatants under the same experimental conditions as for ELISA assays. The presence of the enzymatic activity was detected at 92 kDa corresponding to the pro-MMP-9, as well as at 72 kDa corresponding to the pro-MMP-2 (Figure 1). Pre-incubation of the supernatants with APMA resulted in the appearance of a band at 82 kDa, compatible with the conversion of pro-MMP-9 into the mature form of the enzyme. Treatment of B-CLL cells with HF clearly reduced the gelatinase activity of pro-MMP-9 (92 kDA), whereas an effect on that of the mature enzyme (82 kDa) was hardly detectable; no effect on pro-MMP-2 was observed (Figure 1). These results, in accordance with the above ELISA data, confirm that the secretion of the latent pro-MMP-9 by B-CLL cells is inhibited by HF treatment.

Figure 1

Zymography of MMP-9 gelatinase activity in B-CLL cell supernatants. B-cell lymphocytic leukemia patients' cells (2 × 106/ml) were cultured for 48 h in the presence or absence of 5 μg/ml HF. Cell-free supernatants were collected and further incubated for 60 min at 37°C with or without 0.5 mM APMA, which converts the pro-MMP-9 (92 kDa) into the mature enzyme (82 kDa). Equal volumes of the supernatants were electrophoresed on 7.5% SDS-polyacrylamide gel containing 0.1% gelatin, and gelatinase activity was revealed as described in Materials and methods. Data are from one B-CLL patient representative of three.

Inhibition of matrix metalloproteinase-9 secretion by hyperforin is independent of its apoptotic effect on B-cell lymphocytic leukemia cells

Inasmuch as HF induces caspase-dependent apoptosis of B-CLL cells,18 the decreased MMP-9 production could result from an impaired secretion of dying cells. Although this decreased production occurred while the cells still retained their morphological integrity, the above hypothesis was tested by investigating the effects of z-VAD-fmk, a general inhibitor of caspases. Surprisingly, the presence of z-VAD-fmk significantly reduced MMP-9 relase by untreated control B-CLL cells (Table 2), suggesting a possible MMP-9 regulation through a caspase-dependent mechanism unrelated to apoptosis. In contrast, z-VAD-fmk did not prevent HF-promoted inhibition of MMP-9 secretion, showing that this inhibitory effect of the phloroglucinol is not related to its caspase-dependent apoptotic activity (Table 2).

Table 2 Effects of the pan-caspase inhibitor z-VAD-fmk on MMP-9 release by hyperforin-treated and untreated B-CLL patients' cells

Regulation of matrix metalloproteinase-9 expression and secretion by hyperforin in B-cell lymphocytic leukemia cells

To investigate the mechanisms by which HF regulates MMP-9 production by B-CLL cells, the effect of HF on the number of MMP-9-secreting cells was studied using a specific ELISpot assay. As shown in Table 3, a dose-dependent decrease in the number of secreting cells was detected after treatment of the leukemic cells with HF for 18 h, to reach more than 50% of reduction with 5 μg/ml. Of note, the size of the spots was similar for HF-treated and untreated cells. It appears therefore that HF acts by reducing the number of B-CLL cells releasing MMP-9. Western blot analysis revealed that, in addition, HF decreases expression of pro-MMP-9: indeed, the latent form of 92 kDa, which was strongly expressed in untreated B-CLL, was markedly down-regulated upon HF treatment and this effect was dose-dependent (Figure 2a). Matrix metalloproteinase-9 expression was also studied at the mRNA level by semiquantitative RT-PCR. As illustrated in Figure 2b, MMP-9 mRNA levels were not altered after 18 h of HF treatment, indicating that there is no early regulation of MMP-9 mRNA expression by HF. In contrast, upon prolonged exposure to the phloroglucinol (72 h), a marked dose-dependent inhibition of this expression was observed, suggesting a late MMP-9 mRNA downregulation during HF treatment (Figure 2c).

Table 3 Hyperforin decreases the number of B-CLL cells secreting MMP-9
Figure 2

Effects of HF on MMP-9 protein and mRNA expression in B-CLL cells. B-cell lymphocytic leukemia patients' cells were treated for various times of culture with various concentrations of HF or left untreated (control=C). (a) Western blotting analysis after 48 h of culture: equal amounts of the cell lysates (50 μg of proteins) were immunoprecipitated with an anti-MMP-9 antibody and the immunoprecipitates were analyzed as described in Materials and methods. (b and c) RT-PCR analysis of MMP-9 mRNA as described in Materials and methods, using β2-microglobulin (β2-M) as an internal standard; (b) MMP-9 mRNA expression after treatment of B-CLL cells from two patients with HF for 18 h; (c) MMP-9 mRNA expression after HF treatment of cells from four patients for 72 h. A representative experiment with one patient is shown, and the histograms represent the mean ratio of MMP-9/β2-M mRNA (N=4 patients).

Hyperforin inhibits VEGF but not bFGF release by B-cell lymphocytic leukemia cells

VEGF and bFGF, two regulators of angiogenesis, are secreted by B-CLL cells2, 5, 25, 26 and a neutralizing antibody against VEGF reduced MMP-9 production.5 We therefore investigated the effects of HF on the levels of MMP-9, VEGF and bFGF, as assayed simultaneously in the same culture supernatants of B-CLL cells. Results showed that HF treatment elicited a significant dose-dependent decrease in VEGF production, which closely paralleled its inhibitory effect on MMP-9 release (Figure3a and b); in contrast, bFGF levels remained unaffected by HF (Figure 3c).

Figure 3

Comparative effects of HF on MMP-9, VEGF and bFGF release by B-CLL cells. B-cell lymphocytic leukemia cells from five patients were seeded at 2 × 106/ml and treated in culture for 48 h with 1 and 5 μg/ml HF or left untreated. The concentrations of total MMP-9 (a), VEGF (b) and bFGF (c) were measured simultaneously in the same cell-free culture supernatants using three specific ELISA. Data are expressed as mean levels±s.e. (N=5). Statistical analysis was performed with the Statview software using the paired t-test. *P<0.05; **P<0.001; NS: not significant.

Hyperforin disrupts the formation of tubules by human bone marrow endothelial cells

Based on the fact that some endothelial cells spontaneously organize into capillary-like networks when cultured on Matrigel, we assessed the effect of HF on the vasculogenic potential of HBMECs. In the absence of HF, these cells organized themselves into pseudo-vessels: the formation of microtubules was detected as soon as 3 h of culture, increasing gradually by 6 and 9 h (Figure 4). In the presence of HF, but not of the control vehicle (0.5% ethanol, data not shown), the capacity of HBMECs to proceed to such a self-organization was impaired, as illustrated in Figure 4a. Analysis of the pictures with the Saisam software demonstrated that both the diameter and number of pseudo-capillaries were dose-dependently inhibited by HF (Figure 4b and c). We tested whether this impairment of microtubule formation could be due to HF-promoted apoptosis. We found that treatment of HBMECs with the phloroglucinol altered neither DNA fragmentation (Figure 4d), a well-known nuclear event of apoptosis, nor their viability (not shown). Therefore, the effect of HF on microtubule formation cannot result from apoptotic processes.

Figure 4

Hyperforin inhibits capillary formation by HBMEC on Matrigel without inducing apoptosis. Human bone marrow endothelial cell were cultured for 18 h with medium either alone (control) or containing various HF concentrations or with the vehicle (0.5% ethanol). Cells were then detached, washed, and duplicated aliquots were cultured on Matrigel as described in Materials and methods, in the absence (control) or presence of the respective concentrations of HF. Cell cultures were photographed after 3, 6 and 9 h (a). The diameter (b) and number (c) of microtubules formed at 6 h were estimated with the Microvision Saisam software. The effect of HF on the induction of apoptosis in HBMECs was studied by DNA fragmentation (d), as described in Materials and methods; results are expressed as the percentage of nucleosome enrichment in the cytoplasm of HF-treated versus untreated cells.


We show here for the first time that HF is able to inhibit in vitro three biological events known to play major roles in neo-angiogenesis: the natural phloroglucinol decreases the production of MMP-9 by B-CLL patients' cells, as well as that of VEGF, and reduces the formation of microtubules by endothelial cells derived from bone marrow (HBMEC).

Hyperforin inhibits MMP-9 production by acting on the secretion of the latent pro-enzyme (92 kDa), which is spontaneously produced by B-CLL cells and can be subsequently converted into mature MMP-9 capable of gelatinase activity. The spontaneous release of mature MMP-9 is marginal and nearly not altered by HF. These data suggest that pro-MMP-9 would be first secreted prior to becoming operative locally after enzymatic activation induced by B-CLL cell/stroma interaction. The lack of effect of HF on pro-MMP-2 activity seems of interest, taking into account that the expression of MMP-2 is mostly constitutive, whereas that of MMP-9 is generally inducible and requires a transcriptional activation.

Our observation that the suppressive effect of HF on MMP-9 release is not prevented by the caspase inhibitor z-VAD-fmk demonstrates that it does not result from caspase-dependent apoptotic processes which are induced by the phloroglucinol in B-CLL cells.18 Unexpectedly, z-VAD-fmk is capable of markedly impairing MMP-9 secretion by control unstimulated B-CLL cells. This suggests that MMP-9 expression might be regulated through a mechanism involving a caspase-sensitive event, such as the maturation of a pro-cytokine: indeed IL-1β stimulation of MMP-9 expression in human B-lymphocytes is known to result from caspase-1-induced cleavage of pro-IL-1β. As IL-1β behaves as an autocrine factor in B-CLL cells,27 such an upstream event in MMP-9 regulation could occur in B-CLL cells. Preliminary data from our laboratory also indicate that HF can stimulate caspase-1 activity in the leukemic cells.

Experiments performed to provide informations on the mechanisms through which HF inhibits MMP-9 production show that the phloroglucinol acts by reducing the number of B-CLL cells secreting MMP-9 and by downregulating the expression of pro-MMP-9. At the mRNA level no early regulation is detectable, whereas MMP-9 mRNA expression is inhibited after prolonged exposure to HF. This late effect may be due to the release of secondary mediators leading to the amplification of negative signals, although indirect consequences of late apoptotic processes, such as endonuclease activation, cannot be excluded.

Furthermore, our observation that HF inhibits VEGF production by B-CLL cells also appears of interest. This inhibitory effect parallels that on MMP-9 secretion, whereas bFGF release is unaffected. Whether one of these two effects is responsible for the other has not been elucidated. In ovarian carcinoma cells, activated MMP-9 can directly stimulate VEGF secretion28 in addition to its known ability to favor the release of VEGF trapped in the extracellular matrix. Conversely, the fact that neutralizing anti-VEGF antibodies decreased MMP-9 release by B-CLL cells indicates that VEGF can induce MMP-9 secretion in this leukaemia,5 a result also found in a variety of cell types.4, 29 Although a causal relationship remains to be established, it is thus possible that the observed inhibition of MMP-9 production might result from HF-elicited VEGF downregulation. Indeed, due to the fact that B-CLL cells express VEGF receptors functionally active in angiogenesis,29, 30 the hypothesis of an autocrine loop of VEGF has to be considered in B-CLL,31 and moreover, an autocrine stimulation of VEGFR-2 has been described to induce MMP-9 expression as well as cell growth and migration in some leukaemia.32 Furthermore, elevated MMP-9 levels are also produced in the plasma of adult T-cell leukemia (ATL) patients; they are closely related to high VEGF levels and are significantly associated with strong ATL cell infiltration, suggesting that MMP-9 and VEGF could cooperate in invasion processes.33 The situation is different for B-CLL, inasmuch as there is no correlation between the levels of MMP-9 and VEGF in the serum of B-CLL patients (Molica et al.6 and our unpublished results). Recently, it was reported that the constitutive expression of MMP-9 in B-CLL cells depends on p38 MAP kinase activity, and that the survival of B-CLL cells on stroma is impaired upon inhibition of p38 or MMP-9 activity.34 This indicates that MMP-9, in addition to its effect on angiogenesis, contributes to the well-known survival of leukemic cells in a bone marrow stroma microenvironment. Therefore, the blockade of MMP-9 production could also antagonize the antiapoptotic effect of stromal cells in vivo in B-CLL patients.

Finally, our finding that HF disrupts the formation of microtubules by bone marrow endothelial cells without inducing apoptosis is the first demonstration of the capacity of the phloroglucinol to inhibit the formation of new vessels. Based on this result and on the other present data describing the inhibitory effects of HF on B-CLL cell production of MMP-9 and VEGF, it can be assumed that the phloroglucinol might exert such antiangiogenic effects in vivo: through inhibiting the release of MMP-9 by the leukemic cells, HF would impede their capacity to migrate across the endothelium, inasmuch as MMPs are necessary for endothelial cell migration and tube formation.35 In connection with this assumption and in accordance with our results, a recent report has shown that a stable dicyclohexylammonium salt of HF decreases MMP-9 secretion in different cell lines and reduces inflammatory infiltration, neo-vascularization and size or number of metastases in mice injected with melanoma cells.36

Consequently, through its antiangiogenic properties and its proapoptotic effects on B-CLL cells, HF can be regarded as a compound of great interest for the development of new therapeutic strategies for B-CLL. Our recent observations that HF modulates the activity of P-gp, a member of the ATP-binding cassette (ABC) transporter family involved in multi-drug resistance (manuscript in preparation), has also to be taken into account. All these data highlight that HF deserves further study to evaluate its role as a chemotherapeutic agent and a lead compound for the treatment of B-CLL.


  1. 1

    Ribatti D, Scavelli C, Roccaro AM, Crivellato E, Nico B, Vacca A . Hematopoietic cancer and angiogenesis. Stem Cells Dev 2004; 13: 484–495.

    Article  Google Scholar 

  2. 2

    Kini AR, Kay NE, Peterson LC . Increased bone marrow angiogenesis in B cell chronic lymphocytic leukemia. Leukemia 2000; 14: 1414–1418.

    CAS  Article  Google Scholar 

  3. 3

    Wolowiec D, Wozniak Z, Potoczek S, Ganczarski G, Wrobel T, Kuliczkowski K et al. Bone marrow angiogenesis and proliferation in B-cell chronic lymphocytic leukemia. Anal Quant Cytol Histol 2004; 26: 263–270.

    PubMed  Google Scholar 

  4. 4

    Van den Steen PE, Dubois B, Nelissen I, Rudd PM, Dwek RA, Opdenaker G . Biochemistry and molecular biology of gelatinase B or matrix metalloprotease-9 (MMP-9). Critl Rev Biochem Mol Biol 2002; 37: 375–536.

    CAS  Article  Google Scholar 

  5. 5

    Bauvois B, Dumont J, Mathiot C, Kolb JP . Production of matrix metalloproteinase-9 in early stage B-CLL: suppression by interferons. Leukemia 2002; 16: 791–798.

    CAS  Article  Google Scholar 

  6. 6

    Molica S, Vitelli G, Levato D, Giannarelli D, Vacca A, Cuneo A et al. Increased serum levels of matrix metalloproteinase-9 predict clinical outcome of patients with early B-cell chronic lymphocytic leukaemia. Eur J Haematol 2003; 70: 373–378.

    CAS  Article  Google Scholar 

  7. 7

    Kamiguti AS, Lee ES, Till KJ, Harris RJ, Glenn MA, Lin K et al. The role of matrix metalloproteinase 9 in the pathogenesis of chronic lymphocytic leukaemia. Br J Haematol 2004; 125: 128–140.

    CAS  Article  Google Scholar 

  8. 8

    Barnes J, Anderson LA, Phillipson JD . St John's wort (Hypericum perforatum L.): a review of its chemistry, pharmacology and clinical properties. J Pharm Pharmacol 2001; 53: 583–600.

    CAS  Article  Google Scholar 

  9. 9

    Singer A, Wonnemann M, Muller WE . Hyperforin, a major antidepressant constituent of St John's Wort, inhibits serotonin uptake by elevating free intracellular Na+. J Pharmacol Exp Ther 1999; 290: 1363–1368.

    CAS  PubMed  Google Scholar 

  10. 10

    Treiber K, Singer A, Henke B, Muller WE . Hyperforin activates nonselective cation channels (NSCCs). Br J Pharmacol 2005; 145: 75–83.

    CAS  Article  Google Scholar 

  11. 11

    Schempp CM, Pelz K, Wittmer A, Schopf E, Simon JC . Antibacterial activity of hyperforin from St John's wort, against multiresistant Staphylococcus aureus and gram-positive bacteria. Lancet 1999; 353: 2129.

    CAS  Article  Google Scholar 

  12. 12

    Heilmann J, Winkelmann K, Sticher O . Studies on the antioxidative activity of phloroglucinol derivatives isolated from hypericum species. Planta Med 2003; 69: 202–206.

    CAS  Article  Google Scholar 

  13. 13

    Albert D, Zundorf I, Dingermann T, Muller WE, Steinhilber D, Werz O . Hyperforin is a dual inhibitor of cyclooxygenase-1 and 5-lipoxygenase. Biochem Pharmacol 2002; 64: 1767–1775.

    CAS  Article  Google Scholar 

  14. 14

    Zou L, Harkey MR, Henderson GL . Effects of herbal components on cDNA-expressed cytochrome P450 enzyme catalytic activity. Life Sci 2002; 71: 1579–1589.

    CAS  Article  Google Scholar 

  15. 15

    Schempp CM, Winghofer B, Ludtke R, Simon-Haarhaus B, Schopf E, Simon JC . Topical application of St John's wort (Hypericum perforatum L.) and of its metabolite hyperforin inhibits the allostimulatory capacity of epidermal cells. Br J Dermatol 2000; 142: 979–984.

    CAS  Article  Google Scholar 

  16. 16

    Hostanska K, Reichling J, Bommer S, Weber M, Saller R . Hyperforin a constituent of St John's wort (Hypericum perforatum L.) extract induces apoptosis by triggering activation of caspases and with hypericin synergistically exerts cytotoxicity towards human malignant cell lines. Eur J Pharm Biopharm 2003; 56: 121–132.

    CAS  Article  Google Scholar 

  17. 17

    Schempp CM, Kirkin V, Simon-Haarhaus B, Kersten A, Kiss J, Termeer CC et al. Inhibition of tumour cell growth by hyperforin, a novel anticancer drug from St John's wort that acts by induction of apoptosis. Oncogene 2002; 21: 1242–1250.

    CAS  Article  Google Scholar 

  18. 18

    Quiney C, Billard C, Faussat AM, Salanoubat C, Ensaf A, Naït-Si Y et al. Pro-apoptotic properties of hyperforin in leukemic cells from patients with B-cell chronic lymphocytic leukemia. Leukemia 2006, (in press).

  19. 19

    Zhao H, Dugas N, Mathiot C, Dugas B, Sigaux F, Kolb JP . B-cell chronic lymphocytic leukemia cells express a functional inducible nitric oxide synthase displaying anti-apoptotic activity. Blood 1998; 92: 1031–1043.

    CAS  PubMed  Google Scholar 

  20. 20

    Lehr JE, Pienta KJ . Preferential adhesion of prostate cancer cells to a human bone marrow endothelial cell line. J Natl Cancer Inst 1998; 90: 118–123.

    CAS  Article  Google Scholar 

  21. 21

    Fourneron JD, Naït-Si Y . Effect of eluent pH on HPLC/UV analysis of hyperforin. Phytochem Analysis 2006, (in press).

  22. 22

    Billard C, Kern C, Tang R, Ajchenbaum-Cymbalista F, Kolb JP . Flavopiridol down-regulates the expression of both the inducible NO synthase and p27kip1 in malignant cells from B-cell chronic lymphocytic leukemia. Leukemia 2003; 17: 2438–2443.

    Article  Google Scholar 

  23. 23

    Kern C, Cornuel J, Billard C, Tang R, Rouillard D, Steunou V et al. Involvement of BAFF and APRIL in the resistance to apoptosis of B-CLL through an autocrine pathway. Blood 2004; 103: 679–688.

    CAS  Article  Google Scholar 

  24. 24

    Chen W, Valamanesh F, Mirshahi T, Soria J, Tang R, Agarwal MK et al. Aldosterone signaling modifies capillary formation by human bone marrow endothelial cells. Vasc Pharmacol 2004; 40: 269–277.

    CAS  Article  Google Scholar 

  25. 25

    Aguayo A, Kantarjian H, Manshouri T, Gidel C, Estey E, Thomas D et al. Angiogenesis in acute and chronic leukemias and myelodysplatic syndromes. Blood 2000; 96: 2240–2245.

    CAS  PubMed  PubMed Central  Google Scholar 

  26. 26

    Chen H, Treweeke AT, West DC, Till KJ, Cawley JC, Zuzel M et al. In vitro and in vivo production of vascular endothelial growth factor by chronic lymphocytic leukemia cells. Blood 2000; 96: 3181–3187.

    CAS  PubMed  Google Scholar 

  27. 27

    Uggla C, Aguilar-Santelises M, Rosen A, Mellstedt H, Jondal M . Spontaneous production of interleukin 1 activity by chronic lymphocytic leukemic cells. Blood 1987; 70: 1851–1857.

    CAS  PubMed  Google Scholar 

  28. 28

    Belotti D, Paganoni P, Manenti L, Garofalo A, Marchini S, Taraboletti G et al. Matrix metalloproteinases (MMP9 and MMP2) induce the release of vascular endothelial growth factor (VEGF) by ovarian carcinoma cells: implications for ascites formation. Cancer Res 2003; 63: 5224–5229.

    CAS  PubMed  Google Scholar 

  29. 29

    Hiratsuka S, Nakamura K, Iwai S, Murakami M, Itoh T, Kijima H et al. MMP9 induction by vascular endothelial growth factor receptor-1 is involved in lung-specific metastasis. Cancer Cell 2002; 2: 289–300.

    CAS  Article  Google Scholar 

  30. 30

    Kay NE, Bone ND, Tschumper RC, Howell KH, Geyer SM, Dewald GW et al. B-CLL cells are capable of synthesis and secretion of both pro- and anti-angiogenic molecules. Leukemia 2002; 16: 911–919.

    CAS  Article  Google Scholar 

  31. 31

    Bairey O, Boycov O, Kaganovsky E, Zimra Y, Shaklai M, Rabizadeh E . All three receptors for vascular endothelial growth factor (VEGF) are expressed on B-chronic lymphocytic leukemia (CLL) cells. Leuk Res 2004; 28: 243–248.

    CAS  Article  Google Scholar 

  32. 32

    Dias S, Hattori K, Zhu Z, Heissig B, Choy M, Lane W et al. Autocrine stimulation of VEGFR-2 activates human leukemic cell growth and migration. J Clin Invest 2000; 106: 511–521.

    CAS  Article  Google Scholar 

  33. 33

    Hayashibara T, Yamada Y, Onimaru Y, Tsutsumi C, Nakayama S, Mori N et al. Matrix metalloproteinase-9 and vascular endothelial growth factor: a possible link in adult T-cell leukaemia cell invasion. Br J Haematol 2002; 116: 94–102.

    CAS  Article  Google Scholar 

  34. 34

    Ringshausen I, Dechow T, Schneller F, Weick K, Oelsner M, Peschel C et al. Constitutive activation of the MAPkinase p38 is critical for MMP-9 production and survival of B-CLL cells on bone marrow stromal cells. Leukemia 2004; 18: 1964–1970.

    CAS  Article  Google Scholar 

  35. 35

    Nguyen M, Arkell J, Jackson CJ . Human endothelial gelatinases and angiogenesis. Int J Biochem Cell Biol 2001; 33: 960–970.

    CAS  Article  Google Scholar 

  36. 36

    Dona M, Dell'Aica I, Pezzato E, Sartor L, Calabrese F, Della Barbera M et al. Hyperforin inhibits cancer invasion and metastasis. Cancer Res 2004; 64: 6225–6232.

    CAS  Article  Google Scholar 

Download references


This work was supported by INSERM, Canceropôle Ile-de-France and by a grant from ARC (#3322). CQ was supported by a grant of the Société Française d'Hématologie. We thank the clinicians of the Department of Hematology of the Hôtel-Dieu hospital, and Mrs S Pasco-Dubrulle for technical assistance.

Author information



Corresponding author

Correspondence to J-P Kolb.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Quiney, C., Billard, C., Mirshahi, P. et al. Hyperforin inhibits MMP-9 secretion by B-CLL cells and microtubule formation by endothelial cells. Leukemia 20, 583–589 (2006).

Download citation


  • hyperforin
  • B-CLL
  • MMP-9
  • VEGF
  • angiogenesis
  • endothelial cells

Further reading