Tải bản đầy đủ - 0 (trang)
2 Selection of Biomarkers for Key Cell Types Required in Toxicology

2 Selection of Biomarkers for Key Cell Types Required in Toxicology

Tải bản đầy đủ - 0trang


G.N. Stacey et al.

Table 11.4 Candidate biomarkers for quality control of different hPSC-derived cell models

studied in the SEURAT-1 partner consortium Sr&Tox (http://www.scrtox.eu/)

Cell types

Cardiac cells

Cellular markers

Immunological analysis of cellular

marker expression: Tropomyosin,

Troponin I, Actinin, Atrial Natriuretic

Peptide, Desmin

Costaining of MLC-2a and MLC2v

for determination of subtype cell fate

(atrial/ventricle) and maturation

(atrial occurs first in both subtypes, in

ventricular CMs than coexpressed and

later only MLC-2v)

Analysis of gene expression:

brachyury, Nkx2.5, alpha-cardiac

actin, nppa

Hepatic cells

Analysis of markers/genes

expression: CYP3A4, CYP2B6,

CYP1A1/2, CYP2C9, CYP2C19,

CYP2D6, AFP, ALB, Sox17,


Functional assays

Microscopic evaluation of contracting

cells/areas produced in an efficient

differentiation protocol (Lian et al.


When such 2D cultures are used for

generation of 3D cultures the number

of contracting clusters may be scored

as a proportion of seeded

undifferentiated cells (microscopic

evaluation and field potential recoding)

Generation of action potentials (for

example as measured by microelectrode array)

Sensitivity to channel blockers (for

example as measured by microelectrode array)

Functional hormonal regulation

Urea synthesis

Glycogen uptake

Albumin secretion

Fibrinogen secretion

ATP levels

Glutathione (GSH) levels

Drug-metabolizing cytochrome P450

(CYP450) activities (in particular


Analysis of Albumin synthesis

Phase II activities: Measurement of

activities of glutathione S-transferase

(GST) isoenzymes; Measurement of

activities of UDPglucuronosyltransferase (UGT)


Drug transporter capacity: analysis of

ATP-binding cassette (ABC)

transporter expression and activity;

analysis of solute carrier (SLC)

transporter expression



Ensuring the Quality of Stem Cell-Derived In Vitro Models…


Table 11.4 (continued)

Cell types

Neural cells

Cellular markers

Immunological markers:

Neural stem/progenitor cells:

Sox1, Sox2, Pax6, Nestin

(neuroepithelial stem cells)

Sox2, Pax6, Nestin, 3CB2, BLBP

(radial glia-like stem cells)

Neurons (generic):

β-III-tubulin, MAP2, NF200,


Neuronal subtype-specific

Dopaminergic neurons: Tyrosinehydroxylase (TH), FoxA2, En-1,


Cholinergic neurons: ChAT, VAChT,


Serotonergic neurons: 5HT

GABAergic neurons GABA,


Glutamatergic interneurons:


Layer-specific cortical neurons:

FOXP1, FOXP2, CTIP2, calbindin,

DARPP-32, Tbr1, Tbr2, Satb2, Cux1/2

Sensory neurons: Peripherin, Brn3a

Motoneurons: HB9, SMI-32

Analysis of gene expression/

immunocytochemistry for neural


FoxG1, Otx1/2 (forebrain fate)

Pax2, En1/2, Lmx1Aa/b (midbrain fate)

Gbx2, HoxA2, HoxA4 (hindbrain fate)

(specific markers for ventral and dorsal

identities should/could be implemented

for detailed characterization)

Functional assays

Measurement of neurite and axon

formation and extension

Generation of action potentials (as

generated using micro-electrode


Presence of ion channel activity



G.N. Stacey et al.

Table 11.4 (continued)

Cell types





Cellular markers

Analysis of immunological marker

expression (FACS, IF): K5, K14, K10

Pan-CK antibody (CK14, CK15, CK16

and CK19) performed on the terminally

differentiated keratinocytes as a QC

step immediately prior to setting up for

3D differentiation into 3D epidermis

Analysis of gene expression: K5,

K14, DeltaNP63 (marker of

proliferative keratinocytes) involucrin

(marker of senescent keratinocytes)

Analysis of immunological marker

expression (FACS): CD29, CD44,

CD73, CD105, CD166

Functional assays

Capacity to formed a stratified

epidermis confirmed by histological

H&E staining on the 3D tissue to

identify the presence of Stratum

Basal, Stratum Spinosum, Stratum

Granulosum and Stratum Corneum.

Presence of Stratum Corneum is

critical and tolerances should be set

for its thickness based on user

experience. N.B. too thin giving poor

barrier function and false positive

toxicity and too thick yielding false

negative results

Analysis of the proliferation capacity

in presence of increasing serum

concentration e.g. using Cell Titer


Analysis of cells response to statin

treatment and rescue by mevalonate

There are several diverse neural differentiation protocols for hESCs and hiPSCs,

based on embryoid body formation (Carpenter et al. 2001; Zhang et al. 2001), direct

differentiation into a neural lineage in a suspension cultures (Nat et al. 2007; Schulz

et al. 2004) or in adherent cultures in coated well plates or in co-culture with mouse

stromal cells or astrocytes (Baharvand et al. 2007). These differentiation protocols

include several differentiation- and proliferation-inducing factors such as basic

FGF, EGF, fetal bovine serum, inhibitory protein noggin, retinoic acid, BDNF,

GDNF cAMP etc. as well as conditioned medium. The current neural differentiation

protocols are not equally effective for different hESCs and hiPSCs lines, probably

due to the influence of genetic background or derivation and culture methods.

One of the most important issues in the neuronal/glial differentiation is the

detailed characterization of the terminal differentiated cell populations. It is important to show that markers for pluripotency disappear during neuronal differentiation

and markers specific for neuronal and glial maturation are up regulated. It has also

been shown that in neurosphere cultures the process of neuronal differentiation is

more advanced than in monolayer cultures. Typically neurospheres do not express

the pluripotency marker Oct-4 after 3–6 weeks of differentiation but express

Musashi, Nestin and Pax-6 indicating their neural progenitor nature (Lappalainen

et al. 2010).


Ensuring the Quality of Stem Cell-Derived In Vitro Models…


In order to evaluate the suitability of the differentiated neuronal cell population

for neurotoxicological studies the expression of specific differentiation-related

markers should be characterized. In the case of neuronal differentiation the most

obvious candidate proteins include neurofilament 200 (NF200), Synapsin-I and

synaptophysin for synaptogenesis (Pistollato et al. 2012). Studies of neurotoxicity

in vitro studies should be performed in the presence of glial cells (due to their in vivo

supportive role), therefore it is important to identify and quantify the ratio of neurons to astrocytes/microglia/oligodendrocytes in the terminal cell population.

Astrocytes are usually identified by the expression e.g. of GFAP, oligodendrocytes

by the presence of a marker such as Olig1 and oligodendrocytes are stained by antibodies for OX-42 (Lappalainen et al. 2010). A list of commonly used markers for

each of the neurodevelopmental stages is given in Table 11.5. After selection of a

marker panel, specific quality control methods are needed to establish acceptability

criteria (i.e. acceptable level of expression of cell specific markers at different stages

of cells development and maturation) as well as evaluation of neuronal functionality. Mature neuronal cultures derived from PSC should be proven to be

electro-physiologically active, generating action potentials. One of the commonly

used techniques to characterize action potentials of mature neurons is the multielectrode array (MEA) (Hogberg et al. 2011), which is used as an alternative to the

more classical and challenging ‘patch-clamping’ technique. In this case, specific

quality control metrics for the functional activity and threshold levels for positive

controls need to be defined in order to properly judge the neuronal maturation of an

individual cell preparation. In general, a well-defined set of quality control analyses

should serve as basis for acceptance criteria supporting a reduction of intra- and

inter-laboratory variability of the test system as has been shown in ring-trial neurotoxicity studies based on MEA measurements (Novellino et al. 2011).

Current neuronal differentiation protocols developed for hESCs and hiPSCs usually yield a high percentage of neural precursors (>80 %) (Zhou et al. 2010) and a

significant number of target cells (up to 60 %) (Zeng et al. 2010). hiPSCs also give

an opportunity to create a range of patient genotype-specific or disease-specific

models for neurotoxicity testing. Recently, there have been a number of key developments in neurotoxicological assays, including tests developed that measure the

effects of chemicals on dopaminergic neurons (Zeng et al. 2006). A further test

system has been developed based on measurement of neurite outgrowth providing

automated high-content image analysis and high-throughput screening (Harrill

et al. 2010). Screening using neural progenitors and differentiated neural cells (Han

et al. 2009) has also been established. These test systems could be adapted for the

screening of compounds for neurotoxic effects at different stages of neuronal differentiation. However, in this field one of the main aims is to establish human relevant test systems which can predict the effect of a chemical on the function of

neuronal networks measured by MEA techniques which can be applied to high

throughput/content screening.


G.N. Stacey et al.

Table 11.5 Markers used to discriminate between the different stages of renal development and

expected markers of differentiated target cells


Target cell


stem cells




mesoderm (IM)





Ureteric bud


Podocyte cells


tubular cells

Collecting duct



Nanog 16 (Silva et al. 2009),

Oct4 (POU5F1) (Pan et al.


Brachyury, Mixl1 (Lam et al.


Pax2a, Osr1b, Lhx1b (Dressler

2009; Xia et al. 2013)

Six2, Sall1, Hox11, Eya1

(Dressler 2009), Cited1/Cited

2 (Boyle et al. 2007)

HoxB7, Gfra1, c-Ret (Xia

et al. 2013), Cited4

Synaptopodin (Faul et al.

2007) podocin (Boute et al.

2000; Roselli et al. 2002;

Schwarz et al. 2001), nephrin

(Holthofer et al. 1999;

Holzman et al. 1999; Kestila

et al. 1998; Ruotsalainen et al.

1999), podocalyxin (Horvat

et al. 1986; Kerjaschki et al.

1984), CD2AP (Li et al. 2000;

Shih et al. 2001)

Aquaporin 1 (Nielsen et al.

2002), claudin 2 and 10 (Muto

et al. 2010; Van Itallie et al.

2006; Wilmes et al. 2014),

parathyroid hormone receptor

1 (Stacey et al. 2014b),

organic cation transporter 2

(SLC22A2), MATE1

(SLC47A1) and MATE2-K

(SLC47A2), organic anion

transporter 1 (SLC22A6) and

organic anion transporter 3

(SLC22A8) (Motohashi et al.

2013), glutamyl transferase

(GGT) (Glenner and Folk


aquaporin 2, 3 and 4 (Nielsen

et al. 2002), pendrin

(SLC26A4) (Soleimani 2015)

Additional characteristics

Highly proliferative, can

differentiate into all 3 germ


Precursor of the

intermediate mesoderm

Precursor of the

metanephric mesenchyme

and ureteric bud

Glomerulus and tubular

nephron progenitor

Collecting duct progenitor

Large cell body,

interdigitated foot

processes, VEGF and

prostaglandin secretion

(Jennings et al. 2003)

Cobble stone morphology.

PTH dependent cAMP

induction, low to medium

transepithelial electrical

resistance (TEER), dome

formation and paracellular

water transport (Jennings

et al. 2003; Wilmes et al.


AVP dependent cAMP

induction, very high TEER

(Jennings et al. 2003)

Please note these markers are not necessarily exclusive to the designated cell types, but they discriminate from the other cell types in the table


Some authors show that Pax2 expression persists into MM and UB stages


Osr1 and Lhx1 also expressed in the lateral plate mesoderm


Ensuring the Quality of Stem Cell-Derived In Vitro Models…



Development of Human Hepatic Models Derived from Pluripotent

Stem Cells

Hepatocytes derived from pluripotent stem cells are generally foetal in their phenotype. One major additional problem in order to obtain well differentiated hepatocytes is the fact that freshly isolated hepatocytes rapidly dedifferentiate in most in

vitro systems used and lose their ability to perform basic hepatocyte functions

(Richert et al. 2006; Schwartz et al. 2014) (for typical hepatocyte markers see

Table 11.5). The key differences between stem cell derived hepatocytes and adult

human hepatocytes have hitherto limited the use of stem cells as a source for in vitro

modelling of liver responses.

Several different protocols have been used for the differentiation from stem cells

to hepatocyte like cells. A key component is the differentiation of stem cells to

DE-HEP cells using Activin A (Hay et al. 2008). However, most of these protocols do

not provide cells with a useful phenotype. Some improved phenotype is obtained by

using 3D systems for the differentiation process either in hollow fibre bioreactors

(Sivertsson et al. 2013) or in spheroids (Subramanian et al. 2014), where the expression of key genes encoding e.g. albumin production and drug metabolism is improved.

Toxicity assays have been performed using known hepato-toxins and high content image analysis based determination of toxicity (Sirenko et al. 2014) or ATP

based toxicity assays (Ulvestad et al. 2013; Szkolnicka et al. 2014). The results

show that in some cases the sensitivity for drug toxicity in these stem cell derived

hepatocytes are not too far from those produced by primary hepatocytes cultivated

for 48 h, but still they do not yet provide a robust model of drug induced hepatotoxicity

in vivo in man. To a certain extent the latter extrapolation cannot be possible unless

systems are developed that can be used for assays of chronic drug toxicity, which in

vivo often develops only after 4–12 weeks of development. Furthermore, it is important to include immunologically relevant cells which can mimic the drug induced

idiosyncratic reactions, which in many cases are dependent on the action of specific

HLA class II antigens. Usually stem cell derived hepatocytes are stable only for a

maximum of 2 weeks, hampering the use of such cells for chronic drug induced

hepatotoxicity. However, using a 3D collagen matrix culture (3D clump cultures)

CYP3A4 expression at rather relevant levels have been achieved for 75 days

(Gieseck et al. 2014). The introduction of immune cells and non-parenchymal cells

together with the stem cell derived hepatocytes into 3D in vitro systems would of

course be very challenging. In addition, it would be valuable to use cells derived

from patients susceptible to drug induced liver toxicity in comparison with unaffected controls to highlight potential adverse effects in the liver in vivo. However,

because of the current limited knowledge about differentiation of stem cells into

non parenchymal cells and the lack of a useful hepatocyte phenotype in models

derived from hESC or hiPSC,, it is anticipated that such integrated systems will not

be achieved in the near future.

A key issue in the field of stem cell derived hepatocytes are conditions for better

differentiation and 2D or 3D systems for cultivation of hepatocytes that prevent dedifferentiation. Interesting approaches have been taken where iPSC derived hepatocyte


G.N. Stacey et al.

like cells have been co-cultured with endothelial cells and mesenchymal stem cells on

a pre-solidified matrix forming 3D spheroids (liver buds) in 2 days which are transplanted after 4 days into immunodeficient mice. The resulting human liver tissue has

been found to be highly vascularized with many hepatic functions, although bile ducts

are lacking (Takebe et al. 2013, 2014).

Another interesting approach is the use of small molecules for differentiation of

stem cells to hepatocytes and for proliferation of hepatocytes in vitro as discussed

by Shan et al. (2013). This involves cultivating iPSCs on matrigel supported by

conditioned media from primary mouse embryonic fibroblasts in the presence of

Activin A and growth factors, where after the small molecules were added 21 days

post cultivation for 9 days. The hepatocyte phenotype that these authors achieve is

encouraging for hPSC-based hepatic models, with high expression of several true

hepatic genes. It will be interesting to see if this or similar protocols can be successfully reproduced in other labs and further developed.

In conclusion, we currently have stem cell derived hepatocytes that are relatively

fetal, undifferentiated and not yet of the phenotypic level that can replace human

primary hepatocytes with respect to screening for effects such as drug induced hepatotoxicity. Two novel protocols do indeed generate promising liver functionality of

such cells, but are complicated and labour intensive. We would also like to see these

protocols reproduced in other labs. However, finding of novel key factors for the

differentiation of immature cells into hepatocytes might play an important role for

future development of the field. An interesting aspect in this respect is the simple

overexpression of HNF4α for differentiation of HepaRG cells to generate highly

functional differentiated cells (Chen et al. 2014). Further identification of key

transcription factors and other gene products necessary to activate in the right window of cell differentiation might take this field into a new era.


Development of Human Cardiac Models Derived from Pluripotent

Stem Cells

Recently, the use of hiPSC-derived cardiomyocytes (hiPSC-CM) has increased tremendously in the study of basic cardiac (disease) biology, to assess the effects of

drugs on the heart (efficacy), and to assess possible toxic chemical effects. Recently,

Acimovic and colleagues published a review on the available human pluripotent

stem cell-derived cardiomyocytes as research and therapeutic tools (Acimovic et al.

2014). In the following section, the available methods as described in their review

are shortly discussed.

One of the first protocols describing cardiomyocyte formation from pluripotent

stem cells consists of a co-culture of hESC with mouse visceral-endoderm-like

(END-2) cells (Mummery et al. 2003). END-2 cells secrete factors that have a direct

effect on cardiomyocyte differentiation, such as bone morphogenetic factors (BMPs),

nodal/Activin A, fibroblast growth factors (FGFs) and repressors of the canonical

Wnt/β-catenin signalling pathway (Acimovic et al. 2014). Generally, the efficiency

to produce cardiomyocytes using this original protocol was quite low, but modifica-


Ensuring the Quality of Stem Cell-Derived In Vitro Models…


tions to serum- and insulin-free culture conditions, as well as addition of L-ascorbic

acid increased the cardiomyocyte yield (Passier et al. 2005; Freund et al. 2010).

Cardiomyocytes can also be obtained by culturing stem cells as three dimensional

cell aggregates called embryoid bodies (EBs). To increase the cardiomyocyte yield,

specific growth factors and small molecules can be added. For example, short term

BMP4 treatment can be added to promote mesoderm induction (Zhang et al. 2008)

and also p38 mitogen activated protein kinase (MAPK) inhibition is used to increase

the cardiomyocyte yield (Graichen et al. 2008). Further, the canonical Wnt/β-catenin

signalling pathway has an important role in cardiomyocyte differentiation. For efficient cardiomyocyte differentiation, this pathway should be activated in the early

phase, but inhibited in a later phase of differentiation (Lian et al. 2012). Wnt signalling inhibition increases the efficiency of BMP4-directed cardiac differentiation

(Ren et al. 2011). Also, application of the histone deacetylase inhibitor trichostatin

A has been found to enhance EB-mediated cardiac differentiation (Lim et al. 2013).

Growth factors and small molecules are also applied in monolayer-based cardiac

differentiation protocols. Like with EB-mediated cardiac differentiation, efficient

differentiation results can be obtained if with BMP4-directed cardiac differentiation

Wnt signalling is activated in an early phase and inhibited in a late phase of differentiation (Paige et al. 2010). Further improvements can be obtained when insulin is

removed and FGF2 is added (Uosaki et al. 2011). However, it must be noted that

different differentiation protocols result in different ratios of ventricular and atriallike cardiomyocytes (Acimovic et al. 2014).

During the differentiation steps, signs of immaturity are clearly visible (e.g.

expression of pluripotency and mesenchymal markers, such as stage-specific antigen 1 (SSEA1) and mesoderm posterior 1 (MESP1) (Blin et al. 2010) or cardiac

progenitor markers LIM, homeodomain transcription factor Isl1 (Cai et al. 2003)

and homeobox protein Nkx-2.5 (Stanley et al. 2002). When maintained in culture

for a long time, phenotypic features similar as those of adult cardiomyocytes can be

found (Acimovic et al. 2014; Ivashchenko et al. 2013; Lee et al. 2011; Lundy et al.

2013). It also has been suggested that stimuli like electrical stimulation and mechanical stretching improve maturity and functionality of stem-cell derived cardiomyocytes (Nunes et al. 2013).

For a correct interpretation and reproducibility of experiments, it is of utmost

importance that the stem-cell derived cardiomyocytes used are thoroughly characterized. There are no published consensus guidelines for this, but Mordwinkin and

colleagues recently published a list of criteria that could be used for stem-cell

derived cardiomyocyte characterization (Mordwinkin et al. 2013). These are listed

in Table 11.5.


Development of Human Keratinocyte Models Derived

from Pluripotent Stem Cells

Human keratinocytes and human skin models are key to predicting skin irritation and

may be valuable in predicting other toxic effects in skin. Currently, there are a number of in vitro systems that have regulatory acceptance but these are all derived from


G.N. Stacey et al.

primary tissue. The provision of cell lines that can expand indefinitely and are

capable of robustly and reliably producing stratified epidermis in vitro would be of

value to this area of testing. Pluripotent stem cells have been shown to be capable of

generating dermis and express a range of biomarkers characteristic of the skin (Green

et al. 2003; Guenou et al. 2009; Itoh et al. 2011). These markers include the keratins

5, 10, and 14, as well as involucrin and filaggrin, ITGA6, ITGB4, integrins α6 and

β4, collagen VII and laminin 5 (Guenou et al. 2009; Laustriat et al. 2010; Dinella

et al. 2014). The generation of 3D models of skin have also been reported (Petrova

et al. 2014). Systems are also in development to construct micro-physiological skin

models allowing the interactions between the skin and other organs to be studied

(Guo et al. 2013). These systems are very promising areas of development but as with

other cell types derived from PSCs, to date, they have yet to be developed into reproducible differentiation protocols that can be used across a range of PSC lines.


Development of Human “Mesenchymal” Models Derived

from Pluripotent Stem Cells

Mesenchymal stromal cells (MSCs) have the ability to differentiate into a number

of cell types including adipocytes, chondrocytes, osteocytes and muscle. These cells

have a great therapeutic potential due to their regenerative and immune-regulatory

properties (Augello et al. 2010; Glenn and Whartenby 2014; Sutton and Bonfield

2014), they can also be isolated from a number of tissue sources and minimal criteria to characterise bone marrow derived cell types has been established (Dominici

et al. 2006). However, there is considerable confusion in the literature arising from

poor reporting of MSC cultures and it is important to recognise that this is not one

cell type but probably represents at least three fundamentally different groups.

Primary human MSCs have also been investigated for use in vitro, in acute toxicity

testing and results from these studies indicate that MSCs could be a potential candidate cell type for use in basal cytotoxicity assays (Scanu et al. 2011). However, as

with many primary cells, MSCs suffer from a number of issues including; source of

starting tissue, availability of cells and batch to batch variability. These could potentially be circumvented by using hPSCs to derive MSCs. Indeed there are a number

of reports of the successful generation of MSCs from PSCs both in 2D and 3D

systems (Hematti 2011; Chen et al. 2012; Li et al. 2013; TheinHan et al. 2013; Tang

et al. 2014). Human pluripotent stem cell derived MSCs have been shown to be

comparable to their bone marrow derived counterparts in radiosensitivity assays

(Islam et al. 2015). The similarity between fibroblasts and mesenchymal cells is

often debated and indeed there is a school of thought that suggests that these two

cell types are identical (Haniffa et al. 2009; Hematti 2012). The use of fibroblasts,

derived from human embryonic stem cells, as models for genotoxity has recently

been explored (Vinoth et al. 2014) and the study revealed that these hPSC-derived

fibroblasts were as sensitive to genotoxic challenge as other somatic cell types used

in this assay, Again, as with other cell type specific models being established using

PSCs, the robustness and reproducibility of this system is still a challenge but with

time and effort these issues should be resolved.



Ensuring the Quality of Stem Cell-Derived In Vitro Models…


Development of New In Vitro Stem Cell-Based Cell Models

for Toxicology

Clearly, a range of in vitro tissue cell types are achievable with increasingly better

defined hPSC directed differentiation protocols under development. However, the

stem cell community will need to clearly define the use-cases where stem-cell based

in vitro methods have an added value to other test systems or identify use-cases

where none of the in vitro methods based on non-stem cell based systems can be

used. An area of strong relevance to toxicology which is now advancing in terms of

stem cell-derived models is neuronal 2D and 3D cultures and kidney-derived models. The following section draws on the toxicologically relevant exemplar of kidney

to identify the kinds of scientific considerations needed to progress the early stage

development of new stem cell based models.

The kidneys are vital organs which control the constituents of the blood and

thereby regulate whole body homeostasis. Blood is continuously filtered in the

glomerulus, passes into the renal tubule where essential substances such as sodium,

glucose and amino acids are reabsorbed and waste products and excess substances

are secreted. Due to the multitude of transporting and metabolising systems required

to perform these tasks, cells of the kidney interact with a wide variety of chemicals

entities. This, coupled with its ability to concentrate and metabolise compounds,

makes it susceptible to injury by a wide variety of xenobiotics. The cells of the

nephron exhibit a high degree of physiological, morphological and biochemical heterogeneity (Anonymous 1988; Kriz and Kaissling 2000). These properties determine site-specific sensitivities to xenobiotics. The cells of the glomerulus and the

proximal tubule are the most frequently studied in the context of renal disease and

toxicity due to their critical roles in filtration and reabsorption, respectively. Cultured

renal cells, either primary cultures or immortalised cell, have been extensively

employed in physiological and toxicological studies (Wilmes and Jennings 2014;

Jennings et al. 2008, 2014; Dressler 2006). However, there is now a growing interest

in the utilisation of stem cells to derive renal target cells, mostly for tissue engineering purposes. The use of stem cells is also very attractive for the field of toxicology,

not least due to the fact that cells can be derived from target populations.

The use of pluripotent stem cells to derive renal phenotypes, either from embryonic or somatic sources, brings new challenges. And the major challenge currently

faced is the ability to acquire target cells with the desired phenotype. One strategy

being pursued is to drive pluripotent stem cells through the critical stages in renal

development. The kidney and gonads arise from the intermediate mesoderm which

progress into the primary nephric duct consisting of the pronephros, mesonephros

and metanephros (Sariola 2002). The permanent adult kidney derives from the latter. Within the metanephros the ureteric bud invades the surrounding metanephric

mesenchyme and two-way signalling between these two tissues induces branched

morphogenesis, leading to mature nephron development (Davies 2002). The ureteric bud will finally develop into the collecting duct, whereas the metanephric mesenchyme gives rise to both glomerulus and the cells of the renal tubule (Dressler


G.N. Stacey et al.

2009). As pluripotent stem cells develop into multipotent lineages they lose the

expression of critical pluripotent genes such as Nanog and Oct4 (or POU5F1 in

humans). The loss of the expression of these genes is often used to demonstrate successful progression to multipotent lineages. The intermediate mesoderm expresses

factors such as Pax2, Osr1 and Lhx1, which may or may not be lost as differentiation continues to metanephric mesenchyme and ureteric bud lineages (Xia et al.

2013; Takasato et al. 2014). Different protocols have been used to generate reasonably pure populations of intermediate mesoderm, including sequential addition of

BMP4/FGF2, retinoic acid/activin A/BMP2 (Lam et al. 2014) or by activation of

Wnt signaling with the small molecule agonist CHIR99021 (CHIR) to create

brachyury and Mixl1 positive mesendoderm cells (Araoka et al. 2014; Narayanan

et al. 2013). Several different mixes of developmental growth factors have been

used to differentiate the intermediate mesoderm further into metanephric mesenchyme and ureteric bud cells and even to podocyte and proximal-like phenotypes

(Song et al. 2012; Silva et al. 2009). A list of commonly used markers for each of

the developmental stages and also markers and characteristics of the mature phenotypes, which are present in vivo and maintained in primary culture and some cell

lines are given in Table 11.5.

While there has been great success in the derivation of target renal cells from

pluripotent stem cells, there is still a great deal of work that needs to be done. For

example, most of the protocols developed to-date give mixed populations of cells,

in different differentiation states. Traditional strategies for in vitro toxicity studies

rely on relatively pure cultures of the target cell types. However, probably more

troublesome is the lack of temporal phenotypic stability of the derived cells, which

would be problematic for reproducibility and interpretation of chemical exposures.

However, the field is in its infancy and it is hoped that many of these challenges will

be overcome in the near future. Table 11.5 gives examples of some to the key markers which may be useful in the development and control of stem cell-derived models

of kidney tissue.


General Acceptability Criteria of Stem Cell-Derived

In Vitro Toxicology Assays

In order that in vitro toxicity methods based on stem cells-derived cell or tissue

model (test system) can be considered reliable and relevant with global applicability, they must be reliable and robust showing technical reproducibility between different experimental runs, operators, laboratories and source of equipment and

reagents. A key component in assuring such reliability and standardisation of data

outputs is the use of suitable positive and negative controls. The generation of clear

and unambiguous data, and a clear defined concept on how to use the results in the

context of hazard and risk assessment contexts are of high importance. For new in

vitro toxicity methods based on stem cells as a test system, enough historical data

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

2 Selection of Biomarkers for Key Cell Types Required in Toxicology

Tải bản đầy đủ ngay(0 tr)