CRISPR/Cas9 has a very high efficiency of recombination, it requires very
careful design of sgRNA so that it is possible to avoid off-target effects,
especially when performing multiplexed genome engineering. In addition, there
is the problem of mosaicism. Indeed, not in all embryos is the DNA
double-strand break repaired by homology directed repair, but in many cases,
the break is repaired by non-homologous end-joining (Wang et al., 2013).
these challenges, I believe CRISPR/Cas9 seems very promising for the study of
complex traits and diseases. Indeed, it allows the creation of animal models
carrying multiple mutations in genes associated to the disease that are
detected by Genome Wide Association Studies (GWAS). Furthermore, I think that
CRISPR/Cas9 opens the possibility of studying biological processes directly on
humans by editing hESCs.
Impact of Next-Generation Sequencing (NGS) methods on the use of model
Sequencing (NGS) methods enable the generation of millions of DNA sequence
reads in one run. These innovative methods are changing both the way we do
genetics and our choice of model organisms (Mardis, 2008).
NGS and the
new bioinformatics tools can help identify organisms that are most suitable for
the study of a disorder. Moreover, in the case of complex multifactorial
disorders, even when a model organism is available, the identification and
study of all the genes involved would not be possible without NGS. NGS have
made GWAS possible and have contributed to the identification of new disease
pathways (Bush & Moore, 2012).
schizophrenia was mainly a human disorder that was very difficult, due to its
multifactorial nature, to reproduce in rats and mice. However, recently, NGS
and GWAS have identified different genes involved in Schizophrenia like the
DISC-1 gene and the NRG1 gene (Harrison & Weinberg, 2004). Knock-out mice
have been created for the NRG1 gene and they show many of the symptoms that
characterize this disorder. NRG1 is a membrane protein with different isoforms
which is involved in many different process: neuron migration, synaptic
transmission, white matter regulation and onset of puberty. All these processes
are impaired in schizophrenic (Corfas et al., 2004). The NRG1 homozygous mutant
is not viable because this gene is essential, however, the heterozygous is
still viable and presents some of the main symptoms of schizophrenia. These
mice are, indeed, hyperactive and have hypo-functioning NMDA receptors.
there are some main differences. In humans those homozygous for a single
mutation in the NRG1 gene are still viable and the individuals with this
mutation all develop psychotic symptoms. Additionally, the NRG1-/+ mouse model
does not recapitulate all the symptoms of schizophrenia like depression (Stefansson
et al., 2002). Nevertheless, I think the discovery of this gene is promising
because it enables the creation of models for testing potential drugs. In my
opinion, they are also very useful models to test the interaction of this gene
with environmental factors. Many stress factors are associated with
schizophrenia. Furthermore, this mouse model enables monitoring the expression
of this gene under stress condition. Notably, it was discovered that this gene
is hypomethylated under stress (Rhein et al., 2013).
above, the power of GWAS has revolutionized the way we study the molecular
biology of diseases. Hundreds of genetic variants associated with common human
diseases have been identified, which made it easier to study the molecular
biology of the disease directly without using model organism (Aitman et al.,
2011). Does this mean that in the future there is no need to keep using model
is probably no. Most of the genetic variants that are found have only a very
small effect on phenotype. Many loci need also to be functionally verified.
Moreover, because of the heterogeneous genetic background of human studies, the
statistical analysis is not so powerful as they do not consider the effect of
the environment and of other genetic variants. Additionally, many of these
variants are in non-coding regions making them difficult to study (Wangler et
al., 2017). Therefore, in these cases, I believe there is a need for using
cultures as substitute for model organisms
cells can be grown in culture when provided with the right combination of
growth factors, which is very useful in research. For example, it is possible
to test tumour cells for their sensitivity to cytotoxic reagents to design the
best treatment for patients. Human ESCs were first derived from human
blastocysts in 1998 (Thomson et al., 1998) and are a fundamental tool for
researchers. Scientists can also derive human stem cells from differentiated
cells (Takahashi et al., 2008), called Induced Pluripotent Stem Cells (iPSCs),
which are very promising for regenerative medicine and pharmacology study as
they have the same DNA content of patients (Chien, 2008).
In the 1960
and 1970s research was carried out to understand the physiology and pharmacology
of the central nervous system (CNS). Breakthrough studies about action
potential came from experiments using the giant squid axon (Hodgkin &
Huxley, 1952). However, since 1970s, it was clear that to elucidate the biology
of the synapse at the molecular level it would be necessary to study mammal
nervous cells in culture. Indeed, knocking-out an animal gene that is
homologous to a human gene involved in a certain disease does not necessarily
display the same effects as in humans. The human cerebral cortex is much more
complex than the rodent one. Therefore, creation of in vitro cortical network
is one of the main goal for the study of the human brain. Human ESCs were
successfully differentiated into cerebral cortex stem and progenitor cells
using retinoid signalling in conjunction with inhibition of SMAD signalling,
figure 5 (Shi et al., 2012).
In order to
be good substitutes of model organisms, neurons derived from hESCs need to be
able to fire action potentials. Indeed, the paper showed that cortical neurons
derived by differentiation of human ESCs have normal electrophysiological
properties. Using inhibitors of voltage-gated sodium and potassium channel, it
was possible to demonstrate the presence of these channel by doing whole-cell
patch clamp. Synapses were also identified in differentiated hESCs, figure 6
(Shi et al., 2012).
To sum up,
I believe hESCs and iPSCs combined with CRISPR/Cas9 can overcome the limits
imposed by model organisms and increase our understanding of human biology.
Nonetheless, cell cultures studies have many limitations.
3. The future of model organisms
3.1 Cell culture prospects for the future
with cell culture is the fact that the environmental conditions to which cells
are exposed in vivo are much different from those of cell culture. For example,
the epigenetic profile of cell in cultures is different from their in vivo
version (Nestor et al., 2015). New systems for perfecting cell culture are
being developed. The standard systems used are often 2D with up to 2 cell types,
whereas tissues are mainly 3D structures with several different cell types
(Mason & Price, 2016).
Matrigel-based 3D neural cell culture system has been developed to study
Alzheimer’s Disease (AD). The use of this 3D Matrigel was preferred to normal
2D models because of its similarity to the brain in vivo environment.
Additionally, whereas in 2D models the A? peptides can diffuse out, this system allows
accumulation of the A?
peptides, which are therefore able to aggregate to form the contradistinctive
amyloid plaques of AD. This system was also the first to accumulate aggregated
phosphorylated Tau, another feature of AD (Kim et al., 2015). These findings
are quite remarkable given that no other models before managed to reproduce
both these phenotypic marks of AD together.
believe this system has several limitations. Because Matrigel is produced from
tumour sample, it varies from batch to batch. This makes results more difficult
to recapitulate and scale-up. Additionally, this system lacks microglial cells
which are important regulators of brain inflammation and A? clearance. Moreover, the main brain
regions affected by AD, like the hippocampus, are not represented. This makes
it difficult to use this model to study the progression and diffusion of the
disease in the brain (Kim et al., 2015; Qian et al., 2017).
I think these models are useful to study the early stage of the disease and how
amyloid plaques and neurofibrillary tangles form. However, they do not
represent a reliable model for the study of the disease progression.
Additionally, because of the difficulty in setting up and scaling it, this
technology cannot still be used for conducting high-throughput drug screening.
Furthermore, the absence of an immune and vascular system might bias for these
3.2 Towards humanized animal models
blastocysts complementation enables scientists to create chimeras between two
different species, for example rat-mouse chimeras. This can be achieved by
combining rat pluripotent stem cell (PSCs) with a mouse blastocyst.
Additionally, you can also selectively enrich chimerism in an organ of your
choice. Mouse zygotes that are knock-out for a particular gene can be easily
made with CRISPR/Cas9. For example, you can create mice that are knock-out for
the Pdx1 gene, a gene that is important for the formation of the pancreas. Mice
that are homozygous for this mutation do not survive because they do not have a
pancreas. However, if you complement these homozygous mutant zygotes with rat
PSCs, they can support the development of the pancreas. These experiments are
very promising. It would be nice to be able to grow human organs into animal
species that are more similar to humans to create better model of human
diseases and also to have better systems on which to test drugs. Until now,
there has been an attempt to create chimeras between human and pig. However,
the chimeric ratio was very low and the embryos were not viable (Wu et al.,
In my opinion,
the new tools we have will eventually limit the use of animals as model
systems. This will make experiments more ethical. At the same time, however,
new bioethical challenges will arise. Our ability to create hESCs poses
challenges as some are against the destruction of human embryos. Additionally,
the creations of clones and chimeras reinforce the image of scientists as
people who like to play God. Therefore, it will be very important to educate
the community so that people are aware of the positive potential of these
technologies, and how important is to regulate the use of these tools.
The use of
model organisms was fundamental for understanding several biological processes.
History teaches us how important it is to carefully choose a model organism
that facilitates testing our hypotheses. However, the discovery of new
technologies has changed the way in which we study biology. NGS and
bioinformatics enable us to rapidly sequence the genome of many different
organisms, find genes of interests and mutated sequences. CRISPR/Cas9
facilitated genome editing, enabling scientists to target one or multiple genes
in any organism of interest. When considering the use of model organisms to
study human biology and specifically the progression of diseases, we often
observe that we cannot recapitulate all aspects of a disease in animal models.
Cell culture and, specifically hESC and iPSCs, are eliminating this gap, thus
making “human” the new model organism. However, before we can declare the death
of the model organism, these technologies need to be perfected. Therefore, I
personally think that the parallel use of model organisms together with the new
technologies is the path for success.