CNP, the first potential pharmacological therapy for achondroplasia

Under great expectations, parents and relatives of children bearing the fibroblast growth factor receptor 3 (FGFR3) mutation causing achondroplasia have been following the news about the development of the first real potential drug therapy to treat this condition. In the last quarter of 2010, Biomarin, a pharmaceutical company working in therapies for rare and genetic conditions, announced it was planning to start the clinical research with a compound called BMN-111. BMN-111 was described as a C-type Natriuretic Peptide (CNP) analogue. An analogue is a compound (or molecule) which has a very similar structure compared to the original one, normally keeping the same properties or with enhancements to a given characteristic of that compound.

During 2011, new updates have been released and, in the last International Congress of Human Genetics held in Montreal, a poster describing the results of the BMN-111 tests made in a mouse model of achondroplasia was presented, showing impressive results in terms of the ability of this molecule in restoring the bone growth of those animals: F. Lorget et al. BMN 111, a CNP analogue, promotes skeletal growth and rescues dwarfism in two transgenic mouse models of Fgfr3-related chondrodysplasia.

In December 2011, Biomarin has released new information about the pre-clinical development of BMN-111, presenting results of tests made not only in mice but also in non-human primates, a requisition for any candidate drug to be accepted as an investigational new drug (IND) by regulatory bodies such as the Food and Drug Administration (FDA).

More recently, Biomarin has also announced that it was starting the first clinical trial, a Phase 1 study, to learn how the drug acts in the human body.

Step by step, it seems that the first medicine made to help bones grow in achondroplasia is advancing in its development, an exciting perspective. But, before lighting fireworks, it would be interesting to learn more about CNP and what we should expect about its use in the treatment of children with achondroplasia.

 The Natriuretic Peptide Family

Peptides are molecules made of amino acids, like proteins, but they are smaller and, like their larger cousins, they are also encoded in genes. Natriuretic means a property of something that causes sodium to be eliminated in the urine. The name natriuretic came after the description of one of the first recognized properties of these peptides, which is exactly promoting the elimination of sodium in urine. 

The history of the natriuretic peptides begins about 30 years ago when the first peptide of the family was discovered in extracts of rat atria and for this was called atrial natriuretic peptide (ANP). Not much longer, BNP was identified in extracts of pig brains and then CNP, and as it was the third in the row, received the C-type NP name (this is a link to a good review on CNP by Olney RC). While ANP and BNP are most found in cardiac tissue and are linked to cardiac physiology and related diseases, CNP is expressed in a number of other body tissues and remarkably found within the cartilage growth plate, where it exerts the most important of its actions. This link will take you to a figure showing the three peptides.

CNP

CNP is a known positive player in the growing bone according to many studies made in animal models and also in related mutations in spontaneous human cases. For instance, genetic mutations in the CNP gene causing its overexpression lead to overgrowth. The research also showed that slight changes in the natriuretic peptide receptor type C structure (NPR-C, one NP receptor that is thought to serve as a CNP clearance system) may be responsible for the higher final height found in people from some of the Northern European countries (Estrada K et al.; Bocciardi et al.).

This peptide is expressed (produced) locally in the growth plate. When it binds to its preferential receptor enzyme, NPR-B, located across the cell membrane (in the same way FGFR3 is) of the chondrocyte, it activates this receptor, which in turn causes the activation of other enzymes in the cell cytoplasm. Interestingly, this CNP cascade of chemical reactions will then cross with one of the most important cascades responding to FGFR3 activation, the RAF-RAS-MAPK pathway (discussed here).

When activated by FGFR3, the RAS-RAF-MAPK pathway will lead to one of the most well characterized consequences of achondroplasia, which is slowing down the rate the chondrocytes enlarge (hypertrophy) and mature, thus impairing the entire cartilage growth pace. By the other side, when CNP activates its receptor, the chemical messages emitted by its cascade will turn off or reduce the RAS-RAF-MAPK cascade activity, so causing an inverse action in terms of bone growth. The main observed characteristic of growth plates of mice models of achondroplasia treated with CNP is an enlargement of the hypertrophic zone of the growth plate. Take a look in this article (free access) by Drs Yasoda and Nakao, two of the most prominent researchers of CNP in achondroplasia. They tell the history and results of their research, which has strongly contributed to the understanding of CNP in achondroplasia. This article also has a very didactic graphic showing the two cascades described above.

With the reassuring results of the research by scientists like Drs. Yasoda and Nakao, it became clear that CNP could be used somehow to rescue the bone growth arrest in achondroplasia. But how?

The Japanese group developed a mouse model where the CNP was naturally produced in large quantities by the animal body, in a strategy to simulate a situation where the peptide would be given continuously to the patient. This was necessary because of the nature of CNP. Being a small peptide, it is an usual target of several enzymes present in the blood stream called peptidases. This is so true that CNP, after a single intravenous injection, would last less than five minutes circulating. With such a short half-life (the way scientists describe the interval of time half the quantity of a drug will take to be processed by the body) giving multiple injections would not be a clever strategy to treat any situation. So they probably thought about a therapeutic scheme where their CNP would be given trough a continuous pump infusion, in the same way other clinical conditions have been treated in the past. Their work showed CNP indeed cause bone growth in an achondroplasia +/CNP+ composite mouse model, rescuing the bone growth arrest.

However, this solution, although feasible, has a lot of practical challenges easy to foresee. Then, is there any other way we could give CNP to a child to treat achondroplasia? The answer is yes. Given the strategy announced by Biomarin, in which their CNP analogue will be given subcutaneously once a day, there are other ways. As there is no publicly available information about the compound formula or structure we can only speculate about the solution they found, but it might be related to the knowledge we have about the metabolism of the NPs. Let’s talk a little bit about this.

As mentioned above, CNP and the other related peptides are natural victims of peptidases present in the blood and other tissues. However, the most relevant of these enzymes, neprilysin, does not cleave (cut) the NPs in the same way. Neprilysin has different affinities with the NPs, being ANP and CNP more easily cut than BNP. If you visited the figure I presented above, you may have already identified the structural differences among the three NPs. BNP has two “legs” or branches leaving the main ring while CNP has only one. Evidence exists that the longer BNP branch would be the responsible to its relative resistance to neprilysin (Potter LR, free access). So, there is a chance Biomarin could have developed a CNP analogue bearing a slight modification in its only branch structure (like in BNP) that would give it more resistance to neprilysin activity. This change could give this CNP analogue more time to circulate and diffuse into the tissues and especially into the cartilage growth plate.

It looks like a very smart solution. Tests made with animals have been showing positive results (links above) and, given the FDA authorization to let them proceed to clinical trials, results have been robust enough in terms of efficacy and safety in those animal models.

Testing CNP in clinical trials

Now is the time to test the CNP analogue in humans. What we should expect about these experiments in terms of safety and efficacy?

Safety

First, as CNP is closely related to the other NPs, and that both ANP and BNP have significant effects in the blood pressure and other circulatory parameters, a strict oversight on cardiologic and other circulatory indexes must be carried on. Biomarin has showed during a public presentation in December that the CNP analogue did cause a decrease in the blood pressure in monkeys after each injection.

Second, CNP is found in other tissues throughout the body, including the brain. A recently published study by Dr Nakao and colleagues showed that CNP can influence the body weight, possibly by acting directly in the brain. The mouse model used by the Japanese group does not reproduce the real life, so their results must be understood under this context. Nevertheless, it will be important to follow patients using CNP chronically to understand this aspect of CNP.

Third, bones are not equal, some are thin others thicker. Furthermore, achondroplasia is described as a rhizomelic (rhizo means root) bone dysplasia. This means that it is recognized that proximal (to the trunk) bones are more affected that the distal ones (those in the extremities). There is a theory this could be caused by distinct influences FGFR3 would have across the skeleton, with some bones being more affected than others by the mutation. In some of the papers published by the Dr. Nakao's group, pictures of mice treated with continuous CNP could cause the impression that had thinner spines and longer feet and tails than the control (normal, non-treated) animals. Again, here the kind of exposure those animals had was quite different of what we would expect in the real life or with a single CNP shot a day. Nevertheless, this could be a good aspect to be observed throughout future studies in affected patients.

Fourth, another aspect to be taken in account is the kind of effect the extra CNP would have in other cartilaginous tissues such as the joints, ears, nose and trachea. Although having some specific patterns, chondrocytes tend to behave similarly to the same stimuli wherever they are located, so this is also a question that will need an answer, too.

Efficacy

How will the efficacy of the treatment of CNP be measured? Growth is not a parameter easy to measure in the short term. However, there are some indexes which can be used to monitor the growth rate in children under treatment. For instance the average growth speed rates can be derived from the NCHS series. You can see how this was made by examining this Brazilian Ministry of Health guideline directed to pediatric health care which uses these derived curves (sorry, it is in Portuguese, but look at the page 21 to see the derived graphic).

Growth tends to be fast in the first year after birth and then it starts to slow down up until puberty. It is likely that in children with achondroplasia, taking in account the intrinsic growth impairment, the growth pace could be similar. In fact, Horton and coworkers have showed this in their pivotal study about growth in achondroplasia published in 1978 (J Pediatrics 1978;93 (3):435-8).

The idea could be to plot the already known child heights over the years and create an individual graphic. With the exposure to CNP it would be expected that the growth speed would increase and this can be better measured comparing to the previous pace and to the expected ongoing pace. This is more than just measuring the absolute height.

Another possible marker of growth could be to take measures of the four member bones or, in other words, the lengths of the arms and forearms and thighs and legs could be taken. Then, during the treatment these measures could be readdressed to look for trends in the growth pace in the different member segments. Achondroplasia is a rhizomelic dysplasia, so it would be interesting to learn about the response of the proximal bones to the treatment and also this measure would help to spot earlier any tendency for overgrowth of the extremities.

We must remember that everything in a child with achondroplasia is normal but the workaholic FGFR3. So, if the mutated receptor is compensated what we can expect in terms of bone growth? Doctors know, for a long time, the ‘catch up growth’ phenomenon, seen in several distinct clinical conditions. When the reason for growth impairment is resolved, the affected child tends to grow faster than the average for the age until an individual mark is reached and the growth normalizes. Could the catch-up growth phenomenon happen to children with achondroplasia treated with CNP? This is difficult to say, because in this case the receptor would still be active (so, in a future therapy with a FGFR3 inhibitor, the catch up growth could be expected). Nevertheless, measuring the growth speed would give an insight about this phenomenon in the context of the treatment of achondroplasia with CNP.

The arrival of the CNP analogue as the first potential therapy to help children with achondroplasia to rescue, at least partially, the bone growth, is remarkable. There are several steps to be taken in this phase of its development; the drug must prove to be safe and to have the expected efficacy.  Growing more, affected children could be spared from suffering the many common interventions seen in achondroplasia, from removal of tonsils and adenoids to serious orthopedic and neurological complications. At this moment we must be rational, not presuming that the bone growth will be restored to its full potential. However, in the case of this first possible therapy succeed, a better quality of life could be expected for children with achondroplasia.

O mundo do RNA: desligando o FGFR3 na acondroplasia, parte 2

No último artigo, começamos a rever um complexo controle do funcionamento celular que alguns fragmentos distintos de RNAs exercem. Esses chamados RNAs não-codificantes (ncRNAs) podem regular quanto será criado de uma determinada proteína ao interferir na sua produção, um fenômeno chamado interferência de RNA (RNA interference, RNAi). Os cientistas também chamam essa habilidade de silenciamento de gene (gene silencing), porque impede um gene ativo de dar origem à proteína que ele codifica. Existem várias classes destes RNAs reguladores capazes de interferir na produção de proteínas, tais como os micro RNAs (miRNA) e os pequenos RNAs de interferência (small interfering RNA, siRNA), que descrevemos muito brevemente, que quando acoplados a um complexo de proteínas chamado RISC, ligam-se a um RNA mensageiro (mRNA) e provocam a sua degradação.

RNAs feitos pelo homem

Assim que os pesquisadores começaram a aprender que a máquina genética era muito mais complicada do que os primeiros modelos desenhados lá atrás nas décadas de cinquenta ou sessenta, quando o DNA foi descoberto, eles perceberam que nós poderíamos criar nossas próprias moléculas feitas de código genético, tanto de DNA quanto de RNA.

Recentemente, após a descoberta dos ncRNAs, a introdução das primeiras moléculas artificiais de RNA em células em laboratório e a percepção de que isto poderia ser uma ferramenta útil para entender como funcionam os genes ocorreu de forma ainda mias rápida. O passo seguinte foi uma conclusão natural: se podemos interferir na produção das proteínas (ou no funcionamento dos genes) nestas culturas de células, podemos fazer o mesmo em um tecido vivo? A resposta é novamente sim. Mais do que isto, algumas destas moléculas já estão registradas como medicamentos ou em ensaios clínicos.

Os oligonucleotídeos

Moléculas de RNA artificiais são feitas de blocos de nucleotídeos e a combinação de nucleotídeos em sequências específicas pode ser facilmente executada no laboratório. Em teoria, se alguém quiser bloquear um determinado mRNA, digamos que o mRNA que codifica a proteína receptor do fator de crescimento fibroblasto 3 (FGFR3), é apenas o trabalho de produzir a sequência correta de nucleotídeos que correspondem, ou complementam, a sequência presente no mRNA do FGFR3.

Mas como poderia esta combinação parar a produção da proteína?

Vamos dar uma olhada em uma regra que toda a maquinaria genética segue: A leitura do gene é feita na direção 5' (dizemos cinco (five) prime) para 3' (três (three) prime). Estes números referem-se às posições dos átomos de carbono no anel de açúcar (uma ribose), que faz parte do nucleotídeo. Veja mais sobre isso aqui.

Toda a máquina de transcrição genética e a tradução (ou translação) das proteínas segue esta regra. Nós já vimos um pouco sobre isso no último artigo. No DNA (isto é válido para o RNA, também), temos duas fitas combinadas em uma dupla hélice. Um dos filamentos, que nós chamamos de senso, começa em 5' e termina em 3'. A outra fita, complementar, irá combinar com a cadeia de senso no sentido oposto, a partir de 3' para 5' (sendo chamada de fita antisenso). O que isso tem a ver com RNAi? O efeito dos ncRNAs ocorre pela ligação a uma região próxima à extremidade 3' da sequência do mRNA, o que leva ao bloqueio da leitura do mRNA pelo ribossomo. Convido você a assistir novamente a animação sobre interferência de RNA patrocinada pela revista Nature. Desta vez, preste atenção à posição onde o miRNA se ligará o RNA mensageiro. Há mais de um tipo de interferência de RNA e esta animação mostra os dois tipos principais: a primeira que aparece conduz diretamente à degradação do RNA; a segunda mostra a interferência acontecendo após a ligação ao ribossomo.

Quando os investigadores constroem uma molécula de nucleotídeos olham para a mesma região do alvo natural dos ncRNAs, que está localizada no final da sequência de mRNA sendo, por isso, chamados de oligonucleotídeos antisenso. Oligo é uma raiz para poucos, e dessa forma essa a palavra significa literalmente poucos nucleotídeos. Precisamos de apenas de uma curta sequência de nucleotídeos para fazer a tradução de proteínas parar. No entanto, a investigação sobre a interferência de RNA tem sido muito criativa e não está limitada a este mecanismo natural de ação. De fato, existem algumas terapias antisenso já em estudos clínicos onde o oligonucleotídeo se liga a uma seção particular do mRNA para produzir o efeito desejado. Por exemplo, há um oligonucleotídeo antisenso sendo desenvolvido para o tratamento da distrofia muscular de Duchenne, um tipo muito grave de malformação genética dos músculos, causada por um defeito em uma proteína muito importante chamada distrofina (texto com acesso livre). A molécula funciona ligando-se a uma parte específica do mRNA imaturo da distrofina causando uma mudança na forma como ele é preparado para a tradução, uma estratégia chamada salto de éxon (exon skipping).

Aptâmeros

Se você está acompanhando essa série de artigos, provavelmente vai lembrar dos aptâmeros. Aptâmeros são moléculas muito adaptáveis ​​feitas de nucleotídeos. Mostramos que um aptâmero específico pode ser usado para se ligar ao domínio extracelular de FGFR3 e bloquear a ativação da sinalização do receptor. Bem, a capacidade destes oligonucleotídeos não está limitada a ligação às proteínas. Certamente, eles podem se ligar também a outras sequências de nucleotídeos, o que inclui, potencialmente, o mRNA do FGFR3, através da estratégia de antisenso. Estas duas aplicações potenciais dos aptâmeros na acondroplasia ainda estão esperando por um investigador interessado.

Finalmente, se fazer oligonucleotídeos é tão fácil, então o que estamos esperando? Vamos apenas produzir um oligonucleotídeo para se ligar ao mRNA do FGFR3 e tudo acontece: sem FGFR3, nenhum freio ao crescimento ósseo. Nós já temos exemplos de que isso pode ser feito, veja aqui.

O grande desafio

Existem várias potenciais abordagens para tratar um grande número de condições, incluindo a acondroplasia. O que está nos impedindo de usar um oligonucleotídeo antisenso para parar a produção de uma única proteína, o FGFR3, em um corpo vivo? O principal problema não é o de criar a molécula, mas de fazê-la chegar à célula alvo, o condrócito. Vamos rever o desafio da administração do medicamento no próximo artigo.

Você pode aprender mais sobre interferência de RNA lendo os artigos que mencionei no final do último artigo.

The RNA world: knocking down FGFR3 in achondroplasia, part 2

In the last article, we started to look at a complex cell functioning control some distinct fragments of RNAs exert. These so called non-coding RNAs (ncRNAs) can regulate how much a given protein will be created by interfering in its production, a phenomenon called RNA interference (RNAi). Scientists also call this ability gene silencing, because they hinder an active gene to give origin to the protein it encodes. There are several classes of these regulatory RNAs capable of interfering in the protein production and we have already very briefly described the micro RNAs (miRNA) and the small interfering RNAs (siRNA), which when coupled to complex of proteins called RISC, bind to a messenger RNA (mRNA) and provoke its degradation.

Man-made RNAs

As soon as the researchers started to learn that the genetic machine was far more complicated than the first models designed back in the fifties or sixties, when DNA was discovered, they realized that we could create our own molecules made of genetic code either of DNA or RNA. Recently, following the discovery of the ncRNAs, it was indeed faster to introduce the first artificial RNA molecules into cells in the lab and learn that this could be an useful tool to understand how the genes work. The next step was a simple conclusion: if we can interfere on protein production (or gene functioning) in these cell cultures, can we do the same in a living tissue? The answer is again yes. More than this, some of these molecules are already registered as medicines or in clinical trials.

The Oligonucleotides

Artificial RNA molecules are made of blocks of nucleotides and the combination of nucleotides in certain specific sequences can be easily performed in the lab. In theory, if someone wants to block a given mRNA, let´s say the mRNA which encodes the fibroblast growth factor receptor 3 (FGFR3) protein, it is just the work of combining the right sequence of nucleotides that would match, or complement, the sequence present in the FGFR3 mRNA.

But how could this combination stop the protein production?

Let´s take a look in a rule all genetic machinery follows: the 5’ (we say five prime) to 3’ (three prime) reading direction. These numbers refer to the carbon atom position in the sugar ring (a ribose) which is part of the nucleotide. You can learn more about this here. The entire transcription and translation machinery follows this rule. We have already seen a little about this in the last article. In the DNA (this is valid for RNA, too), we have two strands combined in a double helix. One of the strands, which we call sense strand, starts at 5’ and ends in 3’. The other, or complementary strand, will combine with the sense strand in the opposite direction, from 3’ to 5’ (being called anti-sense strand). What does this have with RNAi? The ncRNAs work by binding to a region close to the 3’ end of the sequence of the mRNA and block the mRNA reading by the ribosome. I invite you to watch again the animation about RNA interference sponsored by the NATURE Journal. This time, pay attention to the position the miRNA will bind the messenger RNA. There is more than one kind of RNA interference and this animation shows the two main kinds: the first one that appears in it is the one which leads directly to RNA degradation; the second shows the interference happening after the ribosome binding to the mRNA.

When researchers build a molecule of nucleotides they look for the same region the natural ncRNAs target and as it is located in the end of the mRNA sequence, they were called anti-sense oligonucleotides. Oligo is a root for few, so the word means literally few nucleotides. We need only a short sequence of nucleotides to make protein translation coming to a stop. Nonetheless, research on RNA interference has been very creative and is not limited to this natural mechanism of action. In fact, there are some antisense therapies already in clinical trials where the molecule binds to a particular section of the mRNA to produce the desired effect. For instance, there is an antisense olgonucleotide being developed to treat Duchenne Muscular Dystrophy, a very severe kind of genetic muscle malformation, caused by a defect in a very important protein called dystrophin (free text). The molecule works by binding a specific part of the dystrophin immature mRNA causing a change in the way it is prepared for translation, a strategy called exon skipping.

Aptamers

If you have been following these articles, you probably will remember the aptamers. Aptamers are very adaptable molecules made of nucleotides. We showed that a specific aptamer could be used to bind the extracellular domain of FGFR3 and block the receptor signaling activation. Well, the ability of these oligonucleotides is not limited to bind proteins. Of course, they can bind also other nucleotide sequences, which includes, potentially, the FGFR3 mRNA, through the antisense strategy. These two potential applications of aptamers in achondroplasia are still waiting for an interested investigator.

Finally, if making oligonucleotides is that easy, then what we are waiting for? Let's just make an oligonucleotide to bind the FGFR3 mRNA and make it happens: no FGFR3, no bone growth arrest. We already have examples that this can be made here. 

The great challenge

A lot of potential approaches to treat a vast number of conditions, including achondroplasia is becoming available. What is preventing us to use an antisense oligonucleotide to stop a single protein, the FGFR3, production inside a living body? The main problem is not about creating the molecule, it is about delivering it to the target cell, the chondrocyte.

We will work on the delivery challenge in the next article.

You can learn more about RNA interference reading the papers I mentioned in the end of the last article.

Biomarin announces phase I clinical trial of BMN-111

Biomarin has just confirmed the initiation of the first study of BMN-111 in humans. This is a phase 1 clinical trial to verify if the CNP analogue is safe, how it is tolerated by the body, what is the path the compound follows in the body (pharmacokinetics) and the optimum dose for the treatment. The following link will take you to Biomarin site, where you can read more details.

http://phx.corporate-ir.net/phoenix.zhtml?c=106657&p=irol-newsArticle&ID=1662147&highlight

Treating achondroplasia

Article originally published in Spanish, English and Portuguese at Fundación Alpe website www.fundacionalpe.org) in September 2011. This text is out of the sequence started in this blog.


Achondroplasia (ACH) is the most common form of dwarfism, with a case estimate of ~1:15000 to 1:25000 births. It is most commonly the result of a de novo mutation of the fibroblast growth factor receptor 3 (FGFR3).

Mutations in the FGFR3 leading to the ACH family of chondrodysplasias are caused by single substitutions of one aminoacid of the protein chain. In the case of ACH, more than 90% of the carriers bear the G380R substitution in the transmembrane domain of the receptor. FGFR3 is mainly expressed in the proliferative chondrocyte region of the growth plate. This receptor tyrosine kinase has a normal negative action on the bone growth and, because the mutation results in a gain of function, the growth plate chondrocytes have an impaired proliferation rate and are induced for an early terminal differentiation, compared to normal individuals, leading to the typical dwarf phenotype.

A note about the cartilage growth plate. This is an avascular, dense, electrically charged environment, functioning as strong barrier for molecules to reach the chondrocytes. Experiments have shown that free diffusion through the growth plate is easier for compounds with less than 50kDa.

What does already exist in terms of pharmacological approaches against FGFR3?

The best strategy should be that directed specifically against the FGFR3, such as antibodies or the tyrosine kinase inhibitors (TKI), but other options are acceptable. Some of these indirect approaches work counteracting the FGFR3 effects in bone growth, although they do not reduce the mutant receptor activity.

• Antibodies anti-FGFR3 have already been developed and show strong and reliable affinity and specificity to the receptor. The main issue is that most of them are large in size (more than 50 kDa), and some tests already performed showed that the dose needed to achieve a therapeutic effect in ACH mice models was toxic and lethal.

• Several TKIs available to date have have shown activity directly against the FGFR3, however none of them seem to be specific enough to allow further development. Most of these TKIs have also action, for instance, against the vascular endothelial growth factor receptor (VEGFR), another receptor enzyme with important actions. The most recently published promising anti-FGFR3 TKI is the NF-449. The NF-449 seems to be more specific than other TKIs but it is still in basic research.

• Parathyroid Hormone (PTH) plays a main role in bone growth. A PTH related protein, PTHrP, is a bone local key factor responsible for maintaining chondrocytes in the proliferation phase, thus positively and strongly influencing the bone growth. Analogues and other compounds to induce sustained release of PTH are being explored for the treatment of osteoporosis and, until recently, there were groups exploring PTHrP actions in ACH models. Some PTH analogues, like teriparatide, are already being marketed to treat osteoporosis in adults. The Food and Drug Administration (FDA) has posed severe restrictions for the use of teriparatide. For instance, its use is forbidden in children, because teriparatide showed to induce osteosarcoma in rats given extremely high doses of this PTH analogue in a life-long manner. Authorities fear that the anabolic effects of teriparatide could lead to a higher risk of cancer development in young, as children have a natural high metabolic rate and would be more prone for neoplastic transformation. It is important to note that, since its launch, there is no one single report linking teriparatide to osteosarcoma in users.

• Menaquinones, represented mainly by menatetrenone (MK-4), are compounds pertaining to the vitamin K2 family. In Japan, MK-4 is used for the treatment and prevention of osteoporosis and also as an adjuvant treatment in advanced liver cancer. A recent paper showed that it acts by reducing the expression of FGFR3 in hepatocarcinoma cells, thus helping to induce them to programmed cell death (apoptosis). It is possible that the same effect in the expression of FGFR3 could be verified in chondrocytes, however this must be tested.

The recommended dose of MK-4 in osteoporosis is 45mg/day (less than 1mg/kg/day in a 50kg adult). This dose is already far larger than the standard nutritional dose recommended by health authorities around the world.

More recently, a Japanese group, looking at long term effects of the use of high doses of MK-4 in rats, found that those rats fed with MK-4 grew more than the control. However, in this paper, the dose of MK-4 the rats were exposed, 30mg/kg/day, is by far higher than the osteoporosis doses and much more indeed than the nutritional ones. Although MK-4 has an excellent safety profile in the recommended dose for the treatment of osteoporosis, a minority of patients suffers of increase in the liver enzymes. Therefore, until specific tests in ACH are performed to check if it can reduce the expression of FGFR3 and also the safety profile is addressed in larger doses, it is too early to say MK-4 or other menaquinone would have a role in ACH.

• C-type Natriuretic Peptide (CNP) is a natural positive player in chondrocyte proliferation by reducing the progression speed of these cells to a more mature state called hypertrophy. It is locally released and seems to not have significant systemic actions, even in large doses. A group from the University of Kyoto, in association with Chugai, a Japanese pharmaceutical industry, has developed a CNP analogue which is frankly in pre-clinical development in mouse models. It seems to rescue most of the features of the ACH phenotype. The caveat is that they are working with a CNP continuous infusion model and this kind of treatment strategy has several practical limitations. To overcome this potential burden, Biomarin, another industry, is developing another CNP analogue (BMN-111) for daily subcutaneous injections. Preliminary results are promising, and Biomarin plans to apply for clinical trials already in 2012.

In summary, currently, the most promising therapeutic option for ACH is the CNP analogue BMN-111, which is planned to start being tested in humans in 2012. Other potential therapies include the blockage of FGFR3 with specific inhibitors, such as NF-449, and the use of PTH or PTHrP to counteract the effects of FGFR3. MK-4 or other menaquinones, must be tested before they should be considered for ACH, due to the very high pharmacological doses supposed to have an effect in FGFR3.

Tratando la acondroplasia

Traducción del original inglés por Fundación Alpe (Gijón), publicado en septiembre de 2011, en www.fundacionalpe.org.

Acondroplasia (ACH) es la forma más común de enanismo, con una estimación de un caso de cada 15000 o 25000 nacimientos. Generalmente es resultado de una mutación de novo del factor receptor de crecimiento de fibroblastos (FGFR3).

Las mutaciones en el FGFR3 que llevan a la familia de condrodisplasias de la ACH son causadas por sustitución simple de un aminoácido de la cadena de proteínas. En el caso de la ACH, más del 90 % de los conductores tienen una sustitución de G380R en el dominio de la transmembrana del receptor. FGFR3 se expresa fundamentalmente en la proliferación de condrocitos en la región de la placa de crecimiento. Este receptor de tirosina quinasa tiene una acción negativa natural en el crecimiento del hueso y, debido a que la mutación conduce a un aumento de la función, los condrocitos de la placa de crecimiento tienen un nivel de proliferación reducido y son inducidos a un diferenciación terminal temprana, en comparación con los individuos normales, que lleva al típico fenotipo del enanismo.

Una nota sobre la placa de crecimiento del cartílago. Este es un entorno sin vascularizaión, denso, eléctricamente cargado, que supone una fuerte barrera para que las moléculas lleguen a los condrocitos. Ciertos experimentos han mostrado que la libre difusión por la placa de crecimiento es más fácil para compuestos con menos de 50 kDa.

¿Qué existe ya, en términos de aproximación farmacológica, contra el FGFR3?

La mejor estrategia debería ser que se dirigiera específicamente contra el FGFR3, como los anticuerpos o los inhibidores de tirosina quinasa (TKI), pero hay más opciones aceptables. Algunas de estas aproximaciones indirectas trabajan contrarrestando los efectos del FGFR3 en el crecimiento del hueso, aunque no reducen la actividad del receptor mutado.

• Anticuerpos anti- FGFR3 ya han sido desarrollados y muestran una afinidad y especificidad hacia el receptor fuertes y de confianza. El principal problema es que la mayoría son de gran tamaño (más de 50 kDa) y algunas pruebas ya llevadas a cabo demostraron que la dosis necesaria para conseguir un efecto terapéutico en ratones modelo con ACH era tóxica y letal.

• Varios inhibidores de tirosina quinasa han mostrado hasta el momento actividad directamente contra el FGFR3, pero ninguno parece ser suficientemente específico para permitir un desarrollo ulterior. La mayoría de estos TKIs también funcionan, por ejemplo, contra el factor receptor de crecimiento endotelial (VEGFR), otra enzima receptora con importantes acciones. El más prometedor TKI anti FGFR3 publicado es el NF-449. El NF -449 parece ser más específico que otros TKIs, pero aún se encuentra en investigación básica.

• La hormona paratiroidea (PTH) juega un papel fundamental en el crecimiento óseo. Una proteína relacionada con la PTH, la PTHrP, es un factor óseo local clave responsable de mantener los condrocitos en la fase de proliferación, influenciando así positiva y poderosamente el crecimiento óseo.

Se están explorando análogos y otros compuestos para inducir la liberación continuada de PTH para el tratamiento de la osteoporosis y hasta hace poco había grupos estudiando la acción de PTHrP en modelos con ACH. Algunos análogos PTH, como la teriparatida, se están comercializando para tratar la osteoporosis en adultos. La FDA (Food and Drug Administration) ha puesto restricciones severas al uso de la teriparatida.

Por ejemplo, su uso está prohibido en niños porque se mostró que la teriparatida induce osteosarcoma en ratas a las que se les dan dosis extremadamente altas de este análogo PTH a lo largo de toda su vida. Las autoridades temen que los efectos anabólicos de la teriparatida pudieran llevar a un riesgo más elevado de desarrollo de cáncer en los jóvenes, dado que los niños tienen por naturaleza un alto índice metabólico y tendrían mayor riesgo de transformación neoplástica. Es importante observar que, desde su lanzamiento, no se ha informado ni de una sola relación entre la teriparatida y el osteocarcoma entre los usuarios.

• Las menaquinonas, representadas fundamentalmente por la menatetrenona (MK- 4), son compuestos relacionados con la familia de la vitamina K2. En Japón la MK-4 se usa para el tratamiento y prevención de la osteoporosis y también como tratamiento complementario en cáncer avanzado de hígado. Un artículo reciente demostraba que actúa reduciendo la expresión del FGFR3 en las células del hepatocarcinoma, ayudando así a inducirlas a la muerte celular programada (apoptosis). Es posible que se pudiera verificar en los condrocitos el mismo efecto en la expresión del FGFR3, aunque esto ha de ser estudiado.

La dosis recomendada de MK-4 en la osteoporosis es de 45 mg/día (menos del 1mg/kg/día en un adulto de 50 kgs). Esta dosis ya es mucho mayor que la dosis estándar nutricional recomendada por las autoridades sanitarias de todo el mundo.

Más recientemente, un grupo japonés, investigando los efectos a largo plazo del uso de altas dosis de MK-4 en ratas, descubrieron que las ratas alimentadas con MK-4 crecían más que el control. De todos modos, en este artículo, la dosis de MK-4 a la que estaban expuestas las ratas, 30 mg/kg/día, es mucho mayor que las dosis de la osteoporosis y muchísimo más que las nutricionales. Aunque el MK-4 tiene un excelente perfil de seguridad en la dosis recomendada para el tratamiento de la osteoporosis, una minoría de pacientes sufren un aumento en las enzimas del hígado. Así pues, hasta que no se lleven a cabo pruebas específicas en ACH para comprobar si puede reducir la expresión de FGFR3, y el perfil de seguridad se pruebe en dosis más altas, es demasiado pronto para decir que la MK-4 u otra menaquinona tenga un papel en la ACH.

• El péptido natriurético del tipo C (CNP) es un agente positivo natural en la proliferación de condrocitos, reduciendo la velocidad de progresión de estas células a un estado más maduro llamado hipertrofia. Se libera localmente y parece no tener acciones sistémicas significativas, ni siquiera en dosis grandes. Un grupo de la universidad de Kyoto, en asociación con Chugai, industria farmacéutica japonesa, ha desarrollado un análogo CNP que está en fase de desarrollo preclínico en modelos de ratón. Parece rescatar la mayoría de los rasgos del fenotipo de la ACH. La advertencia es que están trabajando con un modelo de CNP de infusión continua y este tipo de estrategia de tratamiento tiene varias limitaciones prácticas. Para superar esta carga potencial, Biomarin, otra industria, está desarrollando otro análogo CNP (BMN-111) para inyección subcutánea diaria. Los resultados preliminares son prometedores y Biomarín planea solicitar ensayos clínicos ya en 2012.

En resumen, actualmente, la opción terapéutica más prometedora para la ach es el análogo del CNP BMN-111, que se piensa empezar a probar en humanos en 2012. Otras terapias potenciales incluyen el bloqueo de los inhibidores específicos de FGFR3, tal como el NF-449, y el uso de PTH o PTHrP para contrarrestar los efectos del FGFR3. MK-4 u otras menaquinonas deben ser probadas antes de tomarse en cuenta para la ACH debido a las dosis farmacológicas muy altas que se supone tienen efecto en el FGFR3.

Reducing FGFR3 influence in achondroplasia, part 4

We have been following the fibroblast growth factor receptor 3 (FGFR3) to find where it would be possible to interfere in its activity in order to counteract the negative effects this receptor produces in achondroplasia (ACH).

We have explored the extracellular part (domain) of the receptor, where aptamers could be developed to block the FGFs’ docking site. We crossed the chondrocyte cell membrane to look at what is called the transmembrane domain where short peptides – called transmembrane interceptors – could be created to hinder the dimmer formation.

We also learned about closing the active areas of FGFR3 in its intracellular domain by specific drugs, the tyrosine kinase inhibitors (TKIs). Closing the FGFR3 electric plugs by TKIs will halt the entire electric/chemical signaling cascade. The expected consequence is that the chondrocytes will be able to proliferate more, get mature (hypertrophic) and give space for bone, restoring a normal (or near normal) bone growth.

Now, let’s go downstream into the signaling cascade.

We can think in FGFR3 signaling cascade like a domino chain. When the first piece is pushed against the second, the entire row will follow, one after another in a chain reaction. In our case, FGFR3 is the first piece of the row and, once activated it will trigger other domino pieces, or other enzymes. Now, think what would happen if we took one or two pieces from the middle of the domino chain. The signal initiated by FGFR3 would not reach its targets inside the cell nucleus thus not causing the effects it is responsible for.

In the last decade, researchers have learned about how FGFR3 influences bone growth by blocking what they believed were its signaling cascades. That’s how we know today which enzymes are relevant for FGFR3. Take a look in this Dr. Horton’s review for a figure showing the cascade. Simplifying a lot, we could describe one of the most relevant FGFR3 chains like this:

FGFR3 > FRS2a/SOS/Grb > Ras > Raf > MEK > ERK/p38 > Nucleus

The other important one is shorter:

FGFR3 > STAT1 > Nucleus

Each of the above acronyms represents an enzyme (or, by analogy, a domino piece).

In theory, we could stop the FGFR3 signaling blocking any enzyme located in this domino chain. Tests have been performed with different enzyme inhibition methods and results confirm the relevance of these enzymes for bone growth. For instance, this study showed that ERK1 and ERK2 inhibition restored bone growth and rescued spine stenosis (enlarged the spinal canal) in an animal model. 

The higher the position in the domino chain, the broader would be the effect of this blockage. On the other side, the lower in the domino chain, the more specific would be the effect of taking out a domino piece.

However, inhibiting enzymes downstream of FGFR3 is not that simple. The enzymes that respond to FGFR3 activation react also to several other receptors, including the other FGFRs (reviewed in a previous post). In other words, there is a risk that blocking, for instance, the RAF enzyme, we would face secondary effects due to the blockage of other receptors’ signaling. This certainly would not be good in a growing body of a child. The risk exists because although FGFR3 is produced almost exclusively by chondrocytes, the enzymes of the domino chain are produced by many other cells in other tissues. What would happen if we gave an anti-ERK aiming to treat ACH? What kind of effects in other cells and, by extension, in other tissues and organs of the body could we expect?

In a child with ACH everything is normal, but a single protein (FGFR3) produced by an unique kind of cell (the chondrocyte), located in a very special tissue (the growth plate cartilage). Taking together the current knowledge about FGFR3 and how it functions, it makes more sense to target FGFR3 directly than other enzymes in its signaling pathway.

In summary, we have reviewed in this series of posts where we could interfere in the function of FGFR3 to reduce its activity in a way that could restore, even partially, the bone growth. So far, there are at least three different potential approaches: the aptamers, the transmembrane interceptors and the TKIs. In terms of development status the most promising today are the TKIs.

We have been working on the enzyme FGFR3, from the moment when it is active there in the cell membrane. Is it possible to interfere with the production of FGFR3? In the next post we will visit the chondrocyte nucleus and search for opportunities where we could obstruct the mutant FGFR3 production. If the altered enzyme is not produced, it can’t cause bone growth arrest.