News and Articles on Health. Medical research topics.

For double staining, we first carried out TUNEL and then proceeded to standard immunocytochemistry using anti-ALK antibody. TUNEL was performed using the DeadEnd Fluorometric TUNEL System (Promega) with the following modifications. The NB-39-nu cells seeded on the 24-well plates that were treated with siRNAs were washed with PBS twice and fixed with 4% paraformaldehyde (methanol free) for 25 minutes at 4°C. The cells were rinsed with PBS twice and then permeabilized with 0.2% Triton X-100 solution in PBS for 5 minutes at room temperature. The cells were washed with PBS twice and covered with an equilibration buffer (from the kit) for 10 minutes at room temperature. The equilibration buffer was drained off, and a reaction buffer containing the equilibration buffer, nucleotide mix, and terminal deoxynucleotidyl transferase enzyme was added to the cells and incubated at 37°C for 1 hour, avoiding exposure to light. The cells were incubated for 15 minutes at room temperature with 2× standard saline citrate to stop the reaction. The cells were washed with PBS three times and then stained for ALK using immunofluorescence as follows. The cells were blocked with 2% bovine serum albumin (Boehringer Mannheim) for 30 minutes at room temperature. The blocking solution was drained off, and the cells were incubated with a 1:1000 dilution of αALK for 1 hour at room temperature. The cells were rinsed with PBS three times and incubated with a 1:40 dilution of rhodamine-conjugated goat anti-rabbit secondary antibody (Santa Cruz Biotechnology) for 30 minutes at room temperature. The cells were washed three times with PBS and mounted in glycerol-based 2.5% 1,4-diazabicyclo[2,2,2] octan. Confocal laser scanning analysis was carried out.

DNA Fragmentation Assay
To detect apoptotic DNA cleavage, DNA fragmentation assay was performed using an Apoptotic DNA Ladder kit (Chemicon International, Inc., Temecula, CA). The cells seeded on the 12-well plates that were treated with siRNAs as previously mentioned were collected in 1.5-ml microcentrifuge tubes. The cells were washed with PBS, centrifuged, and lysed with 20 μl of TE lysis buffer. The lysates were incubated with 5 μl of enzyme A (RNase A) at 37°C for 10 minutes and then at 55°C for 30 minutes after the addition of 5 μl of Enzyme B (Proteinase K). Afterward, 5 μl of ammonium acetate solution and 100 μl of absolute ethanol were added, and the samples were kept at −20°C for 10 minutes. The samples were centrifuged, and the pellets were washed with 70% ethanol. Then the DNA pellets were dissolved in 30 μl of DNA suspension buffer. DNA fragmentations were visualized by electrophoresis on 2% agarose gel containing ethidium bromide.

Genomic DNAs derived from neuroblastoma cell lines were obtained from cultured cells as described using the procedure of Perucho et al.29 Samples of 85 neuroblastoma tissues were collected at the Chiba Cancer Center and stored as forms of genomic DNA. The stage criterion was based on the International Neuroblastoma Staging System. Samples of 5 μg of DNA digested by EcoRI were electrophoresed in 0.8% agarose gel and blotted onto nitrocellulose filters (Hybond-N+; Amersham, Piscataway, NJ). The probes for detecting the ALK gene, N-myc gene, and ShcC gene were used in our previous study. The intensities of these signals were measured using a Molecular Imager FxPro (Bio-Rad). This study was approved by the ethical judging committee of the National Cancer Center and the Chiba Cancer Center of Japan.

RNA Interference Technique
Twenty-one-nucleotide double-stranded RNAs were synthesized and purified using Dharmacon Research (Lafayette, CO). To suppress the expression of ALK protein, two different pairs of ALK siRNAs, ALK-siRNA1 and ALK-siRNA2, were obtained. The sequences were 5′-GAGUCUGGCAGUUGACUUCdTdT-3′ for ALK-siRNA1 and 5′-GCUCCGGCGUGCCAAGCAGdTdT-3′ for ALK-siRNA2, corresponding to coding region 153 to 171 and 399 to 417 relative to the first nucleotide of the start codon, respectively. Entire sequences were derived from the sequence of human ALK mRNA (accession no. HSU62540). An siRNA, targeting a sequence in the firefly (Photinus pyralis) luciferase mRNA, was used as a negative control (Dharmacon) (luc-siRNA). We also used a scramble siRNA, Scramble Duplex II (Dharmacon) (s-siRNA) as a mismatch siRNA control in addition to luc-siRNA.

NB-39-nu cells were trypsinized, diluted with growth medium containing 10% fetal calf serum, and transferred to 12-well plates at 6 × 104 cells per well for 24 hours before transfection. The transfection of siRNA was carried out using jetSI (Poly plus transfection). A total of 100 μl of serum-free growth medium and 4 μl of jetSI per well were preincubated for 5 to 10 minutes at room temperature. While the incubation was being performed, 100 μl of serum-free growth medium was mixed with 5 μl of 20 μmol/L siRNA duplex (100 pmol). Total siRNA amounts of 50, 100, and 200 pmol were checked in preliminary experiments to find out 100 pmol is the minimal and optimal amount in this scale of RNAi. The 100 μl of jetSI serum-free medium solution was added to the 100 μl of siRNA duplex solution, gently mixed, and incubated for 30 minutes at room temperature. The growth medium on the cells was removed, and 800 μl of serum-free medium was added to each well. A total of 200 μl of the entire mixture was overlaid onto the cells, and cells were incubated for 4 hours at 37°C in a 5% CO2 incubator. After incubation, 1 ml of medium containing 4% fetal calf serum was added without removing the transfection mixture (final concentration 2%). The cells were assayed 84 hours after transfection. SK-N-MC cells were seeded in 12-well plates at a concentration of 1.3 × 105 cells per well. These were treated with siRNAs in the same way as NB-39-nu and assayed 48 hours after transfection. In the 24-well plate, the cells were seeded at the same concentration as the 12-well plate, and siRNAs and all other reagents were used at half volume. After transfection, the cells were examined under a light microscope every day.

Immunoprecipitation and immunoblotting were performed as described previously. The polyclonal antibodies against the CH1 domains of ShcC (amino acids 306–371) and the anti-ALK antibody (αALK) that was against the cytoplasmic portion (amino acid 1379–1524) of human ALK were prepared as described previously. An anti-phosphotyrosine antibody (4G10) was obtained from UBI. Anti-p44/42 MAPKs, anti-phospho-p44/42 MAPKs, anti-Akt, and anti-phospho-Akt antibodies were purchased from Cell Signaling (Beverly, MA). Anti-EGF receptor (EGFR), anti-Ret, and anti-TrkA antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). In vitro kinase assay for ALK was performed as previously described. Anti-ALK immunoprecipitates were incubated with or without Poly-Glu/Tyr as an exogenous substrate.

Immunocytostaining
For ALK/TOTO-3, immunostaining using anti-ALK antibody was performed at first, and then nuclei were stained using TOTO-3. The cells seeded on the 24-well plates were washed with phosphate-buffered saline (PBS) three times and fixed with 4% paraformaldehyde (methanol free) for 5 minutes at room temperature. The cells were rinsed with PBS twice and then permeabilized with 0.2% Triton X-100 solution in PBS for 10 minutes at room temperature. The cells were blocked with 5% goat serum and 3% bovine serum albumin–Tris-buffered saline for 30 minutes at room temperature. The blocking solution was drained off, and the cells were incubated with a 1:1000 dilution of αALK for 1 hour at room temperature. The cells were rinsed with PBS three times and incubated with a 1:2000 dilution of Alexa fluor (Molecular Probes, Eugene, OR) and 1: 100 dilution of TOTO-3 (Molecular Probes) for 30 minutes at room temperature. The cells were washed three times with PBS and mounted in glycerol-based 2.5% 1,4-diazabicyclo[2,2,2] octan. Confocal laser scanning analysis was carried out. For ALK/TUNEL, we first carried out TUNEL and then proceeded to standard immunocytochemistry using anti-ALK antibody. TUNEL was performed using the DeadEnd Fluorometric TUNEL System (Promega, Madison, WI) with the following modifications. The NB-39-nu cells seeded on the 24-well plates that were treated with siRNAs were washed with PBS twice and fixed with 4% paraformaldehyde (methanol free) for 25 minutes at 4°C. The cells were rinsed with PBS twice and then permeabilized with 0.2% Triton X-100 solution in PBS for 5 minutes at room temperature. The cells were washed with PBS twice and covered with an equilibration buffer (from the kit) for 10 minutes at room temperature. The equilibration buffer was drained off, and a reaction buffer containing the equilibration buffer, nucleotide mix, and terminal deoxynucleotidyl transferase enzyme was added to the cells and incubated at 37°C for 1 hour, avoiding exposure to light. The cells were incubated for 15 minutes at room temperature with 2× standard saline citrate to stop the reaction. The cells were washed with PBS three times and then stained for ALK using immunofluorescence as follows. The cells were blocked with 2% bovine serum albumin (Boehringer Mannheim, Germany) for 30 minutes at room temperature. The blocking solution was drained off, and the cells were incubated with a 1:1000 dilution of αALK for 1 hour at room temperature. The cells were rinsed with PBS three times and incubated with a 1:40 dilution of rhodamine-conjugated goat anti-rabbit secondary antibody (Santa Cruz Biotechnology) for 30 minutes at room temperature. The cells were washed three times with PBS and then mounted and observed in the same manner as that for ALK/TOTO-3.

Pathologists characterize many aspects of breast cancer specimens but rarely pay much attention to the stroma. Whereas some tumors have more stroma than others, desmoplasia is only one of many properties that vary among individual ductal carcinomas. It seems logical that the pathology assessment should identify risk factors associated with rapid tumor progression; however, the pathologist focuses on issues that are more pragmatic: the extent of tumor, classification into histologic subtypes, and identification of properties that define therapeutic intervention, such as expression of estrogen receptor (ER) or HER2/neu. In this issue of The American Journal of Pathology, Conklin et al1 have demonstrated a stromal signature that has a correlation with progression. Although the potential exists for this signature to be used for prognosis, the main value of their correlation is to identify two important areas for further investigation. First, the stromal signature indicates a biologically significant process because it predicts how the tumor will behave. Defining this process is the first step in learning how to control it. Second, the stromal signature may indicate a novel mechanism of tumor progression if it is independent of other prognostic markers.

Multiphoton Microscopy

The new study is based on research from the Keely laboratory and others who used multiphoton microscopy to examine the structure of collagen fibers in tumors. Multiphoton microscopy (or two-photon microscopy) utilizes lasers that generate pulses of light with extremely high photon density in the near-infrared wavelength range of 700 to 1300 nm. For fluorescence imaging, the photon density is so high that near-simultaneous absorption of two photons by a fluorophore can occur, resulting in activation that is equivalent to absorption of a single photon with twice the energy. For example, with the laser tuned to 920 nm, simultaneous absorption of two photons is equivalent to absorption of a single 460-nm photon that can thus activate green fluorescent protein, which has a broad absorption peak around 500 nm. The nonlinear dependence of this effect on the square of photon density makes multiphoton fluorescence microscopy confocal in nature. It is also more efficient than ordinary fluorescence microscopy; there is less bleaching of fluorophores out of the plane of focus and deeper penetration into tissues because of the reduced scattering of near-infrared photons.

To image unstained collagen fibers in the tumor, Conklin et al used second harmonic generation, an additional but less well known advantage of multiphoton imaging. In second harmonic generation, periodic structures that are not centrosymmetriceg, collagen fibers—can act like frequency doubling crystals. In essence, two photons of one frequency enter the fiber, and one photon with exactly half the frequency leaves it. These second harmonic signals are mostly emitted in the same direction as the excitation light, but some are scattered backward and detected by the microscope used to deliver the excitation signal. The second harmonic signal is dependent on the presence of appropriately polarized structures including collagen fibers, myosin, and even microtubules.

The second harmonic signal for collagen fibers is detectable in paraffin sections of formalin-fixed tissue. Thus Conklin et al examined a standard breast cancer tissue microarray. The orientation of the collagen fibers was determined relative to the tumor cell masses to generate what the authors term the “tumor-associated collagen signature” or TACS. Different TACS categories correspond to various ways that the extracellular matrix can be organized with respect to the tumor cells. Originally these categories were defined by studying mammary tumor development in mouse models. TACS-1 arises first in early tumors with increased numbers of curved, apparently relaxed collagen fibers around the tumor. TACS-2 develops as the tumor grows larger. The surrounding fibers become straight and parallel to the surface of the tumor, probably reflecting stretching of the fibers due to the expansion of the tumor. TACS-3 reflects a significant reorganization of the matrix, so that straight matrix fibers now lead directly into the tumor cell mass. In TACS-3, the fibers can act as pathways along which cells can crawl, as has been seen in multiphoton imaging of metastatic tumors.

Neuroblastoma is one of the most common pediatric tumors derived from the sympathoadrenal linage of the neural crest. Tumors found in patients under the age of 1 year are usually favorable and often show spontaneous differentiation and regression. Amplification of the N-myc gene occurs in approximately 25% of neuroblastomas and correlates with the aggressiveness of the disease. In addition to N-myc gene amplification, the expression of various genes has significant correlation with the stage of and prognosis for neuroblastoma. A high level of TrkA expression is predictive of a favorable outcome,20 whereas TrkB is highly expressed in immature neuroblastomas with N-myc amplification. High expression of caspase-1, -3, and -8 is correlated with favorable neuroblastomas. On the other hand, survivin, which suppresses caspase and promotes the cell survival signal, is significantly expressed, and telomerase is activated in unfavorable tumors. There may be a critical difference in the expression of other molecules, including RTKs, in neuroblastoma. A recent paper showed that full-length ALK is detected in almost one-half of the cell lines derived from neuroblastomas and neuroectodermal tumors. We have recently shown using mass-spectrometry analysis that ALK is a major phosphoprotein associated with hyperphosphorylated ShcC in several neuroblastoma cell lines. In these cells, ALK was markedly activated, and it induced the constitutive phosphorylation of ShcC and mitogen-activated protein kinase (MAPK), regardless of stimulation by epidermal growth factor (EGF) or nerve growth factor. These findings strongly suggest that constitutively activated ALK kinase plays a physiological role in the development of neuroblastoma.

In this study, we investigated the biological function of the constitutively activated ALK kinase in neuroblastoma. The RNA interference (RNAi) technique using specific sets of small interfering RNA (siRNA) was induced to inhibit the ALK gene expression in human neuroblastoma cells with or without gene amplification of ALK. The effects of disrupted ALK expression on cell survival or downstream signaling, such as MAPKs or Akt pathways, are examined to understand the biological meaning of ALK amplification in neuroblastoma cells. We also performed Southern blot analysis of primary neuroblastoma tumors from 85 patients to check whether the ALK gene amplification was actually present in neuroblastoma tissues. Furthermore, we sought the ALK gene expression in human neuroblastoma tissues using immunohistochemical analysis.

Cell Culture
Cell lines of human neuroblastoma were maintained in RPMI 1640 supplemented with 10% fetal calf serum (Sigma, St. Louis, MO), penicillin, and streptomycin at 37°C in a humidified 5% CO2 incubator.

Reverse Transcription-Polymerase Chain Reaction (RT-PCR) Analysis
Total RNA was extracted with ISOGEN (Nippongene Japan, Toyama, Japan) from NB-39-nu and SK-N-MC cells. The PCR primer pair 5′-AGGTTCTGGCTGCAGATGGT-3′ and 5′-ACATTGTTCTCTCGAGTGCAGAC-3′ corresponding to the cytoplasmic portion of human ALK was prepared. As much as 0.25 μg of total RNA was reverse transcribed and amplified with the SuperScript One-step RT-PCR with the Platinum Taq kit (Invitrogen Life Technologies, Carlsbad, CA) in a total volume of 50l including 2× reaction mix, 0.2 μmol/L of each primer, and 1 μl of RT/Platinum Taq Mix. Amplification conditions consisted of cDNA synthesis and predenaturation at 50°C for 30 minutes and 94°C for 2 minutes followed by 25 cycles at 94°C for 15 seconds, 58°C for 30 seconds, and 72°C for 45 seconds. A final amplification for 7 minutes at 72°C finished the PCR. The product was separated with 1.2% agarose gel electrophoresis and analyzed using the Quality One System (Bio-Rad, Hercules, CA).

Anaplastic lymphoma kinase (ALK) is a tyrosine kinase receptor originally identified as part of the chimeric nucleophosmin-ALK protein in the t(2;5) chromosomal rearrangement associated with anaplastic large cell lymphoma. We recently demonstrated that the ALK kinase is constitutively activated by gene amplification at the ALK locus in several neuroblastoma cell lines. Forming a stable complex with hyperphosphorylated ShcC, activated ALK modifies the responsiveness of the mitogen-activated protein kinase pathway to growth factors. In the present study, the biological role of activated ALK was examined by suppressing the expression of ALK kinase in neuroblastoma cell lines using an RNA interference technique. The suppression of activated ALK in neuroblastoma cells by RNA interference significantly reduced the phosphorylation of ShcC, mitogen-activated protein kinases, and Akt, inducing rapid apoptosis in the cells. By immunohistochemical analysis, the cytoplasmic expression of ALK was detected in most of the samples of neuroblastoma tissues regardless of the stage of the tumor, whereas significant amplification of ALK was observed in only 1 of 85 cases of human neuroblastoma samples. These data demonstrate the limited frequency of ALK activation in the real progression of neuroblastoma.

Receptor tyrosine kinases (RTKs) play an important role in regulating diverse cellular processes, such as proliferation, differentiation, survival, motility, and malignant transformation. The activation of RTKs typically requires ligand-induced receptor oligomerization, which results in tyrosine autophosphorylation of the receptors at tyrosine residues. By recruiting specific sets of signal transducer molecules in a phosphorylation-dependent manner, each RTK is capable of inducing individual, specific cellular responses. On the other hand, activation of RTKs by either mutations or overexpression is frequently found in various human malignancies.

Anaplastic lymphoma kinase (ALK) is a 200-kd tyrosine kinase encoded by the ALK gene on chromosome 2p23. ALK was first identified as part of an oncogenic fusion tyrosine kinase, nucleophosmin-ALK, which is associated with anaplastic large cell lymphoma. It was also found as a form of fusion protein with a clathrin heavy chain (CTCL) in myofibroblastic tumors. Full-length ALK has the typical structure of an RTK, with a large extracellular domain, a lipophilic transmembrane segment, and a cytoplasmic tyrosine kinase domain. ALK is highly homologous to leukocyte tyrosine kinase (LTK) and is further classified into the insulin receptor superfamily. The LTK gene is mainly expressed in pre-B lymphocytes and neuronal tissues, whereas expression of the normal ALK gene in hematopoietic tissues has not been detected. Instead, it is dominantly expressed in the neural system. In the developing brains of mice, specific expression of ALK was seen in the thalamus, mid-brain, olfactory bulb, and selected cranial regions, as well as the dorsal root, the ganglia of mice, suggesting a specific role in the development of the embryonic nervous system. Currently, however, the function of ALK in adult normal tissue or carcinogenesis remains an open question. Several studies have recently indicated pleiotrophin or midkine as possible ligands for ALK. Although they appeared to induce the functional activation of ALK, it is still unclear whether these molecules are the physiological ligands of ALK.

Thus TACS-3 is a potential marker of a highly invasive tumor, in which the reorganization of the matrix around the tumor apparently reflects a dramatic increase in the ability to spread and metastasize. The results by Conklin et al provide the first retrospective test of this hypothesis using a tissue microarray of 207 breast cancer patients with median clinical follow-up of 6 years. We will focus here on the most successful scoring method, termed Score 1. After each tissue core was divided into 14 areas, the presence of TACS-3 structures was scored separately by three pathologists. A receiver operating curve analysis determined the optimum threshold and demonstrated that detection of only one or two TACS-3 structures in a tissue core was sufficient.

In univariate analysis, TACS-3 scoring showed a significant association with both disease-specific survival and disease-free interval, each with a hazard ratio of 3. To test whether the TACS-3 scoring might provide a novel independent biomarker, the authors then performed a multivariate analysis. A wide range of markers were evaluated including tumor grade, size, patient age, estrogen receptor, progesterone receptor (PR), HER-2, and node status, among others. Within this group, TACS-3 was an independent prognostic marker for both disease-specific survival and disease-free interval, together with PR, ER, node status, and tumor size. Notably, TACS-3 did not correlate with any other markers tested in the analysis. Given these promising results, a classification and regression tree analysis was then performed for predicting 10-year disease-specific survival. Importantly, the TACS-3 score provided valuable prognostic information for patients with large ER-positive tumors. Patients with TACS-3-positive tumors showed a 40% reduction in survival compared with those who had TACS-3-negative tumors.

The TACS-3 scoring shows significant correlation with survival in an important category of breast cancer patients—those with ER-positive tumors that are >1.35 cm. The challenge is to translate this correlation to the pathology laboratory. The use of multiphoton microscopy for research studies of human cancer tissues, including freshly biopsied tissues, has been reported by a number of groups. Our opinion is that this practice is currently impractical for standard histopathology laboratories because of the expense of multiphoton microscopes and the amount of time required for the analysis. However, with the development of inexpensive laser sources, multiphoton microscopy may be accessible in the future. In considering less expensive substitutes, it is not exactly clear how TACS-3 signals detected by multiphoton microscopy correspond to standard histopathology, but there are several options for analyzing stroma. These options include selective demonstration of collagen fibers by Sirius red or reticulum stains or visualization along with cellular detail in H&E and trichrome stains.

Final Reflections

We suggest that two distinct processes exist: TACS-2 reflects stromal formation and organization controlled by the tumor cells, an orderly phase of growth and maturation that follows initial stromal penetration by tumor cells. In contrast, TACS-3 reflects invasion and disruption of preformed stroma by tumor cells. It therefore seems essential to define the histologic correlates of the two-photon TACS-3 detection so that the signatures can be assessed within a standard pathology work-up. The data by Conklin et al1 also imply a correlation with stromal expression of syndecan-1, a cell surface receptor that links extracellular matrix and cytoskeleton. Their description is a bit confusing, because an earlier article from this group characterized tumor cell expression of syndecan-1.18 The localization of syndecan-1 thus requires clarification, because immunohistochemical detection of syndecan-1 could prove to be an easily detected surrogate of TACS-3.

In summary, Conklin et al1 provide an admirable example of studies that bridge mouse cancer models and human diagnostic pathology. However, application of TACS-3 as a histologic or prognostic biomarker requires translation into parameters that fit into the routine pathology work-up. From a biological perspective, TACS-3 may represent a specific type of matrix organization controlled by tumor cells or a result of aggressive invasion. These alternatives define a valuable investigation into fundamental mechanisms of tumor progression.

Thorough review of the recent literature yields very little in terms of in vivo validation of mechanisms of CD44 signaling in HNSCC. This represents a deficiency in the current state of scientific knowledge. Most work to date related to CD44 signaling in HNSCC has involved analysis of in vitro data from HNSCC cell lines or relies on inferences from immunohistochemical analysis of patient tissue specimens. For breast cancer, there are in vivo data on CD44 signaling. Co-expression of CD44 v10 and CD44s through transfection of nonmalignant human breast epithelial cells was shown to promote tumorigenesis in athymic nude mice, but not for nontransfected or vector-only transfected parental cells.50 Furthermore, we can infer the importance of CD44 on HNSCC progression in vivo from studies of CD44 as a cancer stem cell marker in HNSCC, in which CD44-positive enriched HNSCC cells had greater tumorigenicity in nude mice, compared with CD44-negative HNSCC cells. Nonetheless, studies of CD44-mediated migration, metastasis, and chemoresistance or radiation resistance in an in vivo model of HNSCC are currently lacking in the literature.

Conclusion

In summary, an accumulating body of evidence highlights the important role of HA and CD44 signaling in HNSCC progression. In Figure 2 we present a model summarizing our current understanding of the role of HA and CD44 interaction with oncogenic signaling pathways to promote tumor progression and chemoresistance in HNSCC. Subsequent to the HA/CD44 interaction that recruits and forms a CD44-EGFR-LARG multimolecular complex, multiple downstream signaling pathways are activated, and cross-talk among Ras, RhoA, ROK, and PI-3 kinase can occur, further promoting diverse tumor progression behaviors. Understanding HA/CD44-mediated signaling pathways may lead to improved treatment, early detection, and prevention for this deadly disease. Research to date suggests that targeted inhibition of HA/CD44-mediated signaling combined with conventional chemotherapy agents may be an efficacious strategy, one that should be pursued to improve the future treatment of advanced HNSCC.

Hyaluronan-stimulated intracellular Ca2+ mobilization mediates important components of the CD44 signaling pathways. Some of these Ca2+-mediated pathways may be mediated by the Ca2+ binding protein calmodulin. It is known that calmodulin is involved in the activation of several important enzymes, including calcium/calmodulin-dependent protein kinase type II (CaMKII), a ubiquitous serine/threonine protein kinase. In HNSCC cells, CaMKII activation by HA/CD44-mediated Ca2+ mobilization results in the phosphorylation of diverse substrates that promote various cell functions, including motility, cell cycle progression, and proliferation. CaMKII phosphorylates the cytoskeletal protein, filamin. These HA/CD44-mediated effects on CaMKII and filamin lead to cytoskeleton reorganization and promote tumor cell migration. Wang et al49 linked HA/CD44-dependent CaMKII activity to topoisomerase II regulation in HNSCC cells. Topoisomerase II is a critical regulator of DNA topology and function. Hyaluronan treatment promoted CaMKII-dependent topoisomerase II phosphorylation, resulting in enhancement of topoisomerase II activity and decreased cytotoxicity of etoposide (a topoisomerase II poison).These HA/CD44-mediated effects on CaMKII and topoisomerase II activity enhanced tumor cell survival.

Another important effector of RhoA pathway signaling is Rho kinase (ROK). Activated ROK is known to phosphorylate a number of cytoskeletal proteins, such as myosin phosphatase and adducin, that are highly involved in tumor migration and to promote the secretion of MMPs involved in tumor invasion. Torre et al recently showed that HA/CD44 interaction increased ROK activity in HNSCC cells. Hyaluronan also promoted Rho kinase-mediated myosin phosphatase phosphorylation, resulting in enhanced tumor cell migration, and it increased activated MMP-2 and MMP-9 secretion.

RhoA/Ca2+ Signaling-Regulated Chemoresistance

HA/CD44 interaction has been shown to promote resistance to multiple chemotherapeutic agents in HNSCC, including cisplatin, methotrexate, doxorubicin (Adriamycin), and etoposide.38, 40, 49 Several cell signaling mechanisms appear to promote CD44-mediated chemoresistance in HNSCC, including EGFR-related signaling pathways (as already described here).38 Recent work by our groupsuggests that regulation of Ca2+ may also affect chemoresistance (unpublished data). The median inhibitory concentration IC50 for the chemotherapy agent methotrexate in the HNSCC cell line SCC4 is dependent on both the Ca2+ level and the presence of HA. SCC4 cells grown in 1.2 mmol/L Ca2+ medium had greater resistance to methotrexate than cells grown in low-Ca2+ medium, and the IC50 was increased in the presence of HA at both Ca2+ concentrations.

Phospholipase C and RhoA signaling, which mediate intracellular Ca2+ levels, has been shown to play roles in mediating chemoresistance in HNSCC. Wang et al demonstrated that HA-mediated cisplatin, methotrexate, and doxorubicin resistance could be eliminated with inhibition of PLC. Torre et al found that combined ROK and PI-3 kinase inhibition resulted in a synergistic prosurvival effect in the presence of cisplatin. Thus, HA-mediated chemoresistance in HNSCC may involve multiple pathways, including RhoA-mediated Ca2+ signaling.

Ankyrin is a membrane-associated cytoskeletal protein that directly binds CD44.10 This CD44-ankyrin interaction causes cytoskeleton activation. Hyaluronan binding to CD44 promotes colocalization of CD44 and ankyrin in cholesterol-containing lipid rafts, and this colocalization appears to be a key mechanism in regulating HA-mediated cytoskeleton function and tumor cell-specific behaviors (eg, cell survival, growth, and migration). Ankyrin- and CD44-containing lipid rafts have been documented in three different tumor cell lines (breast cancer, ovarian cancer, and HNSCC).10

Up-regulated expression of the ezrin-radixin-moesin (ERM) family of cytoskeletal proteins is seen in HNSCC and is associated with poor prognosis.45 ERM proteins are linkers between membrane molecules such as CD44 and the cytoskeleton. Ezrin is a key regulator of tumor metastasis. CD44 interacts with ERM proteins and with merlin, a related protein. Loss of merlin results in increased HA/CD44-mediated tumorigenesis, and overexpression of merlin diminishes tumor cell growth. In HNSCC, moesin appears to interact with CD44 to degrade the ECM at the invasive front of oral cancers. Both the level of expression and the subcellular localization (cytoplasmic) of ERM proteins are indicators of clinical outcome in HNSCC. Higher expression and cytoplasmic localization of ERM proteins are indicators of poorer survival in HNSCC. Osteopontin, a ligand for CD44, colocalizes with CD44 and ezrin in fibroblasts, metastatic breast cancer cells, and HNSCC.
RhoA-Regulated PLC, ROK, and Ca2+ Signaling and Cytoskeleton Activation

Similar mechanisms for CD44-mediated Ca2+ mobilization pathways to promote tumor migration appear to exist in HNSCC. In HNSCC, Bourguignon et al37 reported that CD44 physically associates in a multimolecular complex with LARG and EGFR. HA/CD44 interaction induces LARG-specific RhoA signaling. LARG molecules isolated from HNSCC cells were found to function as a GDP/GTP exchange factor for RhoGTPases, and the basal rate of bound GTP increased at least 2.4-fold with the addition of HA.

To establish a linkage between HA/CD44-mediated LARG-RhoA signaling and intracellular Ca2+ regulation, one member of the PLC family, PLCε, was isolated from HNSCC cells with GDP- or GTP-loaded forms of RhoA-GST-conjugated beads. PLCε-RhoA interaction was found to be GTP-dependent. LARG-activated RhoA was shown to stimulate both the PLCε-mediated IP3 production and the IP3 receptor-triggered intracellular Ca2+ mobilization in HNSCC cells. HNSCC cells incubated with Fura-2/AM were treated with HA with or without pretreatment with inhibitors of PLC and IP3 receptor; fluorescence spectrophotometry demonstrated a rise in intracellular Ca2+ with HA treatment, but not in the presence of pretreatment with various inhibitors.40 These findings suggest that Ca2+ signaling in HNSCC cells involves both HA/CD44-dependent and RhoA/PLC/IP3 receptor-regulated processes.

One of the key downstream molecular effectors in EGFR-mediated cell survival is PI-3 kinase. PI-3 kinase is capable of catalyzing the conversion of phosphatidylinositol-3,4-bisphosphate (PIP2) to phosphatidylinositol 3,4,5-trisphosphate (PIP3), which subsequently results in the activation of AKT. AKT has 3 subtypes: AKT-1, AKT-2, and AKT-3. HA/CD44 signaling in HNSCC activates all three AKT subtypes. Activated AKT down-regulates the cell’s apoptotic potential, thus promoting increased proliferation and tumor cell survival. Torre et al reported that HA/CD44 interaction increased PI-3 kinase activity in HNSCC cells; HA also promoted PI-3 kinase-mediated AKT phosphorylation, resulting in increased tumor cell survival.
CD44-EGFR Signaling and Chemoresistance

HA/CD44 signaling not only appears to promote EGFR-mediated prosurvival effects, but also leads to increased chemoresistance in HNSCC through EGFR signaling pathways. Cisplatin is the most common anticancer drug used today for the treatment of HNSCC. Wang et al reported that HA can promote cisplatin resistance in several HNSCC cell lines; however, the HA-mediated cisplatin resistance could be abolished with inhibitors of EGFR and MAPK.38 The mechanism of action of cisplatin is thought to involve apoptosis-induced cell death. Because AKT suppresses apoptosis, its activation by PI-3 kinase has been suggested to play a key role in cisplatin resistance. Torre et al39 demonstrated that LY-294002, an inhibitor of PI-3 kinase, was capable of blocking HA-mediated cisplatin resistance. Taken together, these reports suggest that HA/CD44 can promote cisplatin resistance in HNSCC through EGFR signaling pathways.

Interaction of HA and CD44 Promotes Cytoskeleton Activation, Migration, Invasion, and Chemoresistance Pathways in HNSCC

A hallmark of all solid malignancies is the ability to invade and/or metastasize to distant sites. Tumor cells possess altered signaling pathwaysthat lead to cytoskeleton activation and migration. Additionally, tumor cells secrete extracellular factors, such as matrix metalloproteinases, to allow breakdown and invasion of the surrounding ECM. Both HA, which is a major component of the ECM, and its major ligand receptor, CD44, have been studied for their role in promoting tumor progression properties, including migration and invasion. The mechanisms of how HA and CD44 interact to promote cytoskeleton activation, migration, and invasion were initially elucidated in several cancer and noncancer models.

RhoA signaling appears to be a critical pathway through which HA/CD44 interaction mediates cytoskeleton activation. One of several important mechanisms used by RhoA in the regulation of cellular functions is through alteration of intracellular Ca2+ levels. Phospholipase C (PLC) is a key mediator in intracellular Ca2+ mobilization. When activated by RhoA, PLCs first hydrolyze PIP2 into inositol trisphosphate (IP3), resulting in Ca2+ release from intracellular stores. This Ca2+ release promotes various cell functions, including cytoskeleton activation, cell cycle progression, and proliferation. Another key RhoA pathway effector is Rho kinase (ROK), which has been shown to regulate several cytoskeletal proteins (such as myosin light chain phosphatase) that are highly involved in tumor migration and to promote the secretion of MMPs, which degrade the ECM during tumor invasion.

HA/CD44 interaction has been shown to be tightly coupled with intracellular Ca2+ mobilization pathways and with RhoA pathways in many different cells. CD44 and RhoA are physically associated in metastatic breast cancer cells, and ROK was found to play a key role in CD44-ankyrin interaction and in RhoA-mediated breast cancer oncogenic signaling. In endothelial cells, HA treatment promoted CD44 interaction with ROK, leading to IP3 receptor-mediated Ca2+ mobilization and migration.9 In keratinocytes, HA-mediated Ca2+ mobilization promoted cortactin-cytoskeleton function, leading to adhesion and differentiation.44 CD44 is linked to ankyrin and has been shown to activate MMP-9 during active migration processes in metastatic breast cancer cells.