Caratterizzazione neurofisiologica, neuropatologica e comportamentale della neuropatia periferica da Bortezomib in moldelli murini

In the last 20 years several in vivo rat models of peripheral neuropathy induced by antineoplastic drugs clarified the phatophysiological mechanisms involved in peripheral neurotoxicity. Moreover these models allowed to identify neuroprotection strategies for preventing sensory nerve and dorsal root ganglia (DRG) neurons damages. Only a few cancer cell lines is able to induce the development of cancer in immunocompetent rats, therefore these models do not represent the best way to investigate, at the same time, the antineoplastic activity and the neurotoxic effects of these drugs; by contrast, murine models are widely employed in oncological studies and can be useful for this purpose. During the Ph.D Programme in Neuroscience, the aim of this projecct has been the neurophysiologic, histophatologic and behavioural characterization of 4 murine models of peripheral neuropathy induced by the proteasome inhibitor bortezomib. The neurotoxic side effects of several antineoplastic agents are among the main reason for tretament modification or their limitation as antitumor therapy. The mechanisms underline the onset of peripheral neurophaties are still not clear, even though it is accepted that the neurotoxicity is dose- and drug-dependent. Bortezomib is a proteasome inhibitor and it is used as an efficient chemotherapeutic agent in the tretament of multiple myeloma. In 2003 the Food and Drug Administration (FDA) approved the use of bortezomib for the treatment of advanced multiple myeloma, then in 2008 the drug was promoted as a first-line chemotherapeutic agent for the treatment of myeloma. However patients treated with bortezomib develop a peripheral neuropathy, often combined with neuropathic pain. This project consist of 3 consecutive phases: (A.1) The development of 4 in vivo murine models of bortezomib-induced peripheral neuropathy. Aim: to evaluate the neuropathologic mechanisms involved in bortezomib-induced peripheral neuropathy in two immunocompetent murine strains (Balb/c and CD1) and in two immunodeficient murine strains (Hola Hsd NuNu and Scid). _Aim of work and summary 12 (A.2) The development of a multiple myeloma murine model in association with bortezomib treatment. Aim: to evaluate, at the same time, the antineoplastic activity and the neurotoxic effect of bortezomib in Scid mice. (B) The localization of Ca2+ transporters Calbindin and Parvalbumin in DRG neurons (respectively markers of small size neurons and big size neurons). Aim: verify the distribution of different DRG neurons subpopulations (small size, medium size and big size neurons) to quantify the damage induced by the treatment. Female Balb/c, CD1, Hola Hsd NuNu and Scid mice (initial body weight 19-21 gr.) were treated with bortezomib to develop and to study the neuropathologic mechanisms of the peripheral neuropathy (A.1). Balb/c mice were administered with bortezomib twice a week for 4 weeks at the dose of 0,8 mg/Kg; CD1 and Hola Hsd NuNu were treated with bortezomib 0,8 mg/Kg twice a week for 6 weeks while Scid mice were administered with bortezomib 1 mg/Kg once a week for 5 weeks. During the treatment period we evaluated both general toxicity parameters, like mortality and body weight changes, and neurotoxicity parameters, including the sensory nerve conduction velocity (NCV) in the caudal and digital nerve, the morphological and morphometrical analysis of DRG neurons and sciatic nerve and the assessment of neuropathic pain by behavioural dynamic test. Bortezomib has been well tolerated in all 4 strains: Balb/c and Scid mice treated with the compound showed a statistically significant decrease in body weight compared to the control ones (respectively of 10,65 and 8,17%); on the contrary CD1 mice showed an increase in body weight even though not comparable to the control ones (reduction of 7,9% respect to control mice) and Hola Hsd NuNu mice revealed an increase in body weight similar to control ones. All 4 strains disclosed different neurophisiological alterations. Bortezomib caused a clear reduction both in caudal and in digital NCV. Balb/c showed a decrease in caudal NCV of 27,3% and in digital NCV of 17,9% (compared to control ones). CD1 mice showed a reduction in caudal NCV of 20,3% and in digital NCV 26,9%; Hola Hsd NuNu mice bared a decrease in caudal NCV of 15,2% and in digital NCV of 13,5%; Scid mice showed a reduction in caudal NCV of 21,35% and in digital NCV of 23,83%. _Aim of work and summary 13 The morphological analysis highlighted alterations of DRG neurons and satellite cells induced by bortezomib; moreover the compound caused a severe and frequent degeneration of myelinic fibers. The morphometrical analysis of DRG neurons showed the increase in size of soma, nucleus and nucleolus in Balb/c treated mice respect to control ones. In CD1 mice only the nucleolus was increased in size, while in Hola Hsd NuNu and Scid mice bortezomib caused an increase in both of soma and nucleolus size. Furthermore Bortezomib induced a statistically significant increase of g-ratio (the ratio of the inner axonal diameter to the total outer diameter) and a trend decrease of myelin thickness. Dynamic test highlighted the onset of mechanical allodynia (a pain due to a stimulus which does not normally provoke pain). CD1 e Scid mice showed a statistically significant reduction of the answer to the mechanical stimulus respect to control ones (respectively of 29,1% and 22,1%); on the contrary in Balb/c mice and Hola Hsd NuNu bortezomib caused a not statistically significant decrease of answer (respectively 21,5% and 20,4%). Female Scid mice (initial body weight 19-21 gr.) were injected with 1x106 RPMI 8226 cells (human cellular line of multiple myeloma) and half of these animals were treated with bortezomib 1 mg/Kg once a week for 5 weeks (A.2) to evaluate, at the same time, the antineoplastic activity and the neurotoxic effect of the compound. Also in this phase, parameters of general toxicity and of neurotoxicity were evaluated. Moreover we monitored the tumor increase. Bortezomib has been well tolerated in all groups: mice injected with RPMI 8226 cells revealed an increase in body weight similar to control ones, while mice injected with RPMI 8226 cells and treated with the compound showed a decrease in body weight between the second and the fourth administration that recovered until the end of treatment. Two groups disclosed different neurophisiological alterations. Bortezomib caused a relevant decrease both in caudal (27,1% respect to control animals) and in digital NCV (19,1% respect to control animals) while multiple myeloma reduced only the digital NCV (14,21% respect to control animals). The morphological analysis showed alterations of DRG neurons and satellite cells induced by multiple myeloma that were confirmed in animals treated with bortezomib; moreover multiple myeloma caused a frequent degeneration of myelinic fibers that increased in animals treated with bortezomib. _Aim of work and summary 14 The morphometrical analysis of DRG neurons showed the decrease in size of soma, nucleus and nucleolus in mice with multiple myeloma, while bortezomib caused an increase in size of soma and nucleolus. The morphometrical analysis of sciatic nerve showed the involvement of multiple myeloma in the increase of g-ratio and in the decrease of myelin thickness. The treatment with bortezomib highlighted these results. Dynamic test demonstrated the onset of mechanical allodynia only in mice with multiple myeloma treated with bortezomib: the decrease in the answer to the stimulus is statistically significant in comparison to control ones (14,53%) but it is not relevant if compared to animals with multiple myeloma (9,6%). Furthermore, the evaluation of tumor size demonstrated that bortezomib has an efficient antineoplastic role starting from the first week of treatment. Since the morphometrical analysis of DRG neurons showed the increase in size of soma, nucleus and nucleolus of animals treated with bortezomib, we decided to measure the damage induced by the compound proving the distribution of different neuronal subpopulations (small, medium and big neurons) (B). Unexpectedly Calbindin and Parvalbumin labelled neurons with similar somatic areas both in control animals (respectively 478,1 ± 210,5 µm2 e 451,5 ± 193.5 µm2) and in treated mice (respectively 368,2 ± 144,1 µm2 e 342,6 ± 136 µm2). A difference between the labelling of control animals respect to treated ones is clear, particularly if we consider the percentage of parvalbumin labelling: indeed the DRG neurons of control animals treated represent the 33,15% of all cells while the DRG neurons of treated animals are the 15,36% of all neurons. The characterization of in vivo murine models of bortezomib-induced peripheral neuropathy of this study is relevant since it could represent a starting point for evaluating the anticancer activity of the compound and, at the same time, its neurotoxicity. Furthermore these models could be useful to test the neuroprotective action of new experimental agents and to evaluate the non-interference between the chemotherapeutic agent and the neuroprotectant agents.