|Home | About | Journals | Submit | Contact Us | Français|
Grafting of neural stem cells (NSCs) or GABA-ergic progenitor cells (GPCs) into the hippocampus could offer an alternative therapy to hippocampal resection in patients with drug-resistant chronic epilepsy, which afflicts >30% of temporal lobe epilepsy (TLE) cases. Multipotent, self-renewing NSCs could be expanded from multiple regions of the developing and adult brain, human embryonic stem cells (hESCs), and human induced pluripotent stem cells (hiPSCs). On the other hand, GPCs could be generated from the medial and lateral ganglionic eminences of the embryonic brain and from hESCs and hiPSCs. To provide comprehensive methodologies involved in testing the efficacy of transplantation of NSCs and GPCs in a rat model of chronic TLE, NSCs derived from the rat medial ganglionic eminence (MGE) and MGE-like GPCs derived from hiPSCs are taken as examples in this unit. The topics comprise description of the required materials, reagents and equipment, methods for obtaining rat MGE-NSCs and hiPSC-derived MGE-like GPCs in culture, generation of chronically epileptic rats, intrahippocampal grafting procedure, post-grafting evaluation of the effects of grafts on spontaneous recurrent seizures and cognitive and mood impairments, analyses of the yield and the fate of graft-derived cells, and the effects of grafts on the host hippocampus.
Chronic temporal lobe epilepsy (TLE) is characterized by recurrent partial complex seizures, memory impairments, depression, and substantial decline in hippocampal neurogenesis (Astur et al., 2002; Hattiangady et al., 2004, 2011; Detour et al., 2005; Coras et al., 2010; Hattiangady and Shetty, 2010; Shetty, 2011). Antiepileptic drug therapy, though widely used for controlling seizures, has no effect on the course of the disease and fails to restrain seizures in >30% of TLE patients (Fisher et al., 1998; Strine et al., 2005). Intracerebral transplantation of NSCs or γ-amino butyric acid (GABA) positive progenitor cells (GPCs) is evolving as an attractive therapy for promoting regeneration and repair in various brain disorders including TLE (Shetty and Bates, 2015). Fascination for using NSCs are linked to their properties such as multipotency and ability for self-renewal and the ease by which they can be obtained from multiple regions of the developing and adult brain, human embryonic stem cells (hESCs), and human induced pluripotent stem cells (hiPSCs) (Shetty and Hattiangady, 2007; Hattiangady and Shetty, 2012; Shetty, 2014). On the other hand, interest in utilizing GPCs stems from the advent of novel directed differentiation methods to obtain them in large numbers from hESCs and hiPSCs (Liu et al., 2013).
Studies in neurological disease models have shown that NSCs can survive intracerebral grafting, engraft into the injured brain areas, release a multitude of neurotrophic factors, positively influence the survival of host cells and tissues, and promote functional recovery. Similarly, GPC grafting has shown considerable promise for alleviating deficits in prototypes of several neurological disorders (Shetty and Bates, 2015). Transplantation of apt NSCs in TLE may considerably restrain spontaneous recurrent seizures (SRS) because of their ability to give rise to significant numbers of neurons synthesizing the inhibitory neurotransmitter GABA and/or astrocytes synthesizing the anticonvulsant protein the glial cell line–derived neurotrophic factor (GDNF) ((Waldau et al., 2010; Hattiangady and Shetty, 2012; Hattiangady et al., 2015). Grafting of NSCs may also improve cognitive function in chronic TLE because of their ability to engraft into neurogenic regions of the dentate gyrus and thereby influence the extent of hippocampal neurogenesis. On the other hand, GPC grafting has been shown to reduce seizures through addition of new GABA-ergic interneurons and improved GABA-ergic neurotransmission in the hippocampus of brains afflicted with TLE (Hunt et al., 2013; Hattiangady et al., 2013; Henderson et al., 2014; Cunningham et al., 2014).
In this unit, to provide a detailed methodology involved in evaluating the usefulness of transplantation of NSCs and GPCs in a rat model of chronic TLE, we describe the protocol for grafting NSCs expanded from the MGE of the embryonic day-14 rat fetuses and MGE-like GPCs derived from hiPSCs into hippocampi of rats exhibiting chronic TLE. The protocols mainly include description of the required materials, reagents and equipment, methods for obtaining rat MGE-NSCs and hiPSC-derived MGE-like GPCs in culture, generation of chronically epileptic rats (CERs), intrahippocampal grafting procedure, post-grafting evaluation of the effects of grafts on spontaneous recurrent seizures and cognitive and mood impairments, analyses of the yield and the fate of graft-derived cells, and the effects of grafts on the host hippocampus.
Figure 1 shows a schematic representation of MGE-NSC grafting experiments described in this manuscript.
NOTE: All protocols using live animal studies must be first reviewed and approved by the Institutional Animal Care and Use Committee (IACUC). The experimenter must strictly follow all the guidelines recommended by the IACUC while performing the experiments in animal models.
In this protocol, we describe how to generate rats exhibiting chronic temporal lobe epilepsy characterized by SRS and cognitive and mood dysfunction using a chemoconvulsant chemical [i.e., kainic acid (KA)] to induce status epilepticus (SE). As generation of rats exhibiting chronic TLE requires a time frame of 3 to 5 months, the experiments to be performed on chronically epileptic rats need to be planned well in advance. Furthermore, as the extent of SRS varies between animals (Rao et al., 2006a, 2007; Waldau et al., 2010; Hattiangady et al., 2011), having a larger pool of rats exhibiting chronic TLE would help in choosing animals exhibiting a similar extent (frequency and intensity) of SRS for the transplantation study.
It will be difficult to score stage I to II seizures, which are characterized by salivation, excessive grooming behavior, mastication, wet dog shakes, etc. However, stages III to V seizures are much easier to follow and can be scored using a modified Racine scale (Racine, 1972; Ben-Ari, 1985; Hellier et al., 1998; Rao et al., 2006a).
Following SE, there will be a silent period of 1 to 2 months during which no or only occasional SRS are observed. Therefore, commencing the measurement of behavioral SRS in the 3rd month after SE is ideal. In most of our studies, we intermittently score the frequency and duration of stage III to V SRS at 3 to 6 months post-SE (i.e., 8 hr/week; 4 hr/session; 2 sessions/week, total 32 hr/month) to determine the extent/pattern of chronic epilepsy. From the recorded seizures, calculate the following parameters for every month of observation: the frequency of all (stages III to V) SRS, the frequency of stage V seizures (the most severe form of SRS), the average duration of individual SRS (i.e., the total amount of time spent in seizures/the total number of seizures), and the percentage of time spent in SRS (i.e., the total amount of time spent in seizures/the total duration of observation × 100).
Intermittent scoring of the frequency and intensity of SRS for several months (i.e., 8 hr/week; 4 hr/session; 2 sessions/week, total 32 hr/month) has been found to be sufficient for determining the extent/intensity of chronic epilepsy typified by SRS in male F344 rats (Rao et al., 2006a, 2007; Waldau et al., 2010; Hattiangady et al., 2011). However, measuring the frequency and intensity of SRS with additional hours of observation or continuous (24/7) video monitoring will be superior and especially important if animals mostly exhibit SRS in clusters.
Select groups of age-matched CERs exhibiting a similar extent of SRS (in terms of both frequency and intensity) from a larger pool of CERs for transplantation studies.
The extent of SRS can vary between epileptic animals (Rao et al., 2006a, 2007; Waldau et al., 2010). Choosing animals exhibiting a similar extent of SRS for different experimental groups would facilitate the comparison of changes in the seizure frequency and intensity with specific treatment such as grafting of NSCs or GPCs, sham-grafting surgery alone, epileptic rats receiving cyclosporine alone, or epileptic rats receiving neither surgery nor grafting (i.e. epilepsy alone rats).
In order to facilitate the assessment of improvement in hippocampus-dependent cognitive or memory function with NSC grafting, it is important to examine the extent of cognitive or memory dysfunction prior to cell grafting in the chosen CERs. While one can assess hippocampal-dependent cognitive or memory function using quite a few tests, we have selected object location test (OLT, a hippocampus-dependent memory test) as an example for assessing object location memory function in this article. The different aspects of this test are described below.
This test is performed for assessing the cognitive ability of rats to detect subtle changes in the environment. Maintenance of this function depends upon the integrity of the hippocampus circuitry.
This test is performed in an open field box measuring 100 cm (L) × 100 cm (W) × 60 cm (H). Each animal is subjected to three successive trials in this test (Fig. 3).
A day before the test, all animals need to be handled. Additionally, make sure that all chosen CERs explore the open field apparatus individually for 10–15 minutes. This pre-test exploration of the apparatus reduces anxiety in CERs on the day of testing. This was evidenced through increased time spent in exploration of objects by CERS in trials 2 and 3.
Export data such as times spent with specific objects and the total object exploration time from the software. Calculate percentages of the total object exploration time spent with the object moved to a novel location vis-à-vis the object that remained in its original location. Compare these values within each animal group using two-tailed, unpaired Student’s t-test to determine the ability of animals for place discrimination. The choice to explore the object displaced to a novel location reflects the ability of animal to discern minor changes in the location of objects in its immediate environment. Typically, CERs (not receiving any therapeutic treatment) display a clear impairment in this cognitive function, as they do not show affinity for the object moved to a novel place in trial 3. Rather, they spend either nearly equal amounts of time with the object in the familiar place (FP object) and the object in the novel place (NP object) or spend greater amount of time with the FP object.
Chronic epilepsy is associated with both loss of functional inhibition and a reduction in the number of GABA-ergic interneurons in the hippocampus. Therefore, transplantation of cells that are capable of differentiating into functional GABA-ergic interneurons in the hippocampus may be useful for improving inhibition as well as restraining SRS. In this context, NSCs derived from the embryonic MGE appear ideal as donor cells for grafting because a significant fraction of these cells can differentiate into GABA-ergic interneurons following grafting.
In order to ascertain the differentiation potential of the chosen donor NSCs, it will be useful to examine their differentiation in culture. The protocols we typically use are described below.
For the generation of MGE-like GABA-ergic progenitor cells from hiPSCs (i.e. human GPCs), we employ a protocol developed by Su-Chun Zhang laboratory (Liu et al., 2013) with some modifications.
Thaw the required amount of BD Matrigel™ on ice (see manufacturer’s instructions for dilution). Dilute the matrigel using cold DMEM/F-12 into a 15 ml tube, mix well. Do not allow formation of a gel inside the tube. Immediately transfer 1 ml of the diluted matrigel into each well of a 6-well plate. Uniformly spread the matrigel into the entire area of the well through slow swirling of the plate. Incubate the plate at room temperature (15 – 25°C) for an hr prior to plating cells. Do not allow the matrigel to dry. Remove the excess matrigel after an hour of incubation and immediately add 2 ml of E8 medium to each well.
Thaw all supplements to room temperature. Prepare the complete TeSR™-E8™ medium (Basal Medium + 20× Supplement + 500× Supplement) under sterile conditions. Aliquot the medium into sterile 50 ml tubes and store at −20°C up to six months or at 2 – 8°C for up to 2 weeks.
Thaw 20× supplement to room temperature. Prepare the complete TeSR™-E6™ medium (Basal Medium + 20× Supplement) under sterile conditions. Aliquot the medium into sterile 50 ml tubes and store at −20°C up to a month or at 2 – 8°C for up to 2 weeks.
Dissolve this enzyme in Dulbecco’s Phosphate-Buffered Saline (DPBS) without calcium and magnesium to 10 mg/ml. Dilute this solution further with DPBS without calcium and magnesium to a final concentration of 1U/mL for the use.
Dissolve 20 mg of heparin in 1 ml of DMEM/F-12 medium. Prepare aliquots and store them at −80 °C for up to 6 months.
For 500 ml of NIM preparation, combine 490 ml of DMEM/F-12, 5 ml of NEAA, 5 ml of N-2 supplement and 50 μl of heparin inside a sterile biosafety cabinet. Store the medium at 2–8 °C for up to 2 weeks.
Completely dissolve 5 mg of purmorphamine in 480 μl of ethanol and 480 μl of DMSO. Aliquot the solution and store them at −80 °C for up to 6 months.
Dissolve 500 μg of Shh into 1 ml of sterile DPBS with 0.1% (wt/vol) human serum albumin or BSA for a 500-μg ml−1 stock. Aliquots and store at −80 °C for up to 6 months.
For 50 ml of NDM preparation, combine 49 ml of Neurobasal medium, 0.5 ml of NEAA and 0.5 ml of N-2 supplement inside a sterile biosafety cabinet. Store the medium at 2–8 °C for up to 2 weeks.
Dissolve 100 μg into 1 ml of sterile DPBS with 0.1% (wt/vol) human serum albumin or BSA. Aliquot and store them at −80 °C for up to 6 months.
Dissolve 4.914 mg of cAMP in 10 ml of sterilized water. Aliquot and store at −80 °C for up to 6 months.
From day 24 onwards, neurospheres containing MGE-like GPCs can be used for characterizing and confirming the presence of NKX2.1+ cells through immunostaining procedures, growing long term cultures to generate mature GABA-ergic interneurons or for transplantation studies. For all these purposes neurospheres are broken down to single cells or into smaller clumps containing <10 cells.
Reconstitute the final pellet in a 30–50 μl differentiation medium and count the viable and dead cells using a trypan blue exclusion test. Adjust the density of viable cells to 80,000–100,000 cells/μl.
NOTE: If viability is <75%, wash cells again to obtain the higher ratio of viable cells in the final cell suspension. Typically, the cell suspension exhibiting >80% viability is ideal for transplantation studies.
Figure 5 illustrates the sequence of steps involved in grafting of MGE-like GABA-ergic cells into the hippocampus of rats with chronic TLE.
Prior to the surgery, choose CERs exhibiting a similar extent of SRS as well as cognitive dysfunction. Choosing rats displaying the above characteristics of TLE would facilitate testing the efficacy of NSC grafts for both restraining seizures and reversing the cognitive dysfunction.
NOTE: Animals chosen for sham-grafting surgery need to undergo a surgical procedure that is identical to what is described here for the transplantation of MGE-NSCs or GPCs. The only difference is that these animals will receive a sterile culture medium (1 μl/site) in place of the cell suspension. Furthermore, animals receiving human GPC grafts need to undergo moderate immune suppression to prevent the rejection of grafts through lymphocytic infiltration. This is typically accomplished through daily subcutaneous injections of cyclosporine (at a dose of 10–12 mg/Kg). Injections need to commence 1–2 days prior to grafting and continue during the entire post-grafting survival period. This strategy prevents the rejection of human cell grafts in the rat brain.
NOTE: Personnel protective equipment including sterile surgical latex gloves, disposable lab coats, head cover, booties, surgical masks and approved respirators (for handling human GPCs) should be worn.
NOTE: A glass bead sterilizer should be available to sterilize the stainless steel surgical instruments during the surgery of multiple animals in one surgery session.
Analyses of rat MGE-NSC or human GPC Grafting-Mediated Changes in Seizures.
In addition to the quantification of behavioral SRS as described above, one can also quantify the long-term changes in all SRS (i.e., electrographic seizures with or without a behavioral component) after the grafting using continuous electroencephalographic (EEG) recordings for 2–6 weeks at an extended time-point after grafting.
It is ideal to do EEG recordings following completion of all behavioral measures of SRS and cognitive and mood function.
In our laboratory, we use a time-locked video-digital EEG monitoring system (AS40 from Grass Telefactor) for measuring SRS from CERs.
For EEG recordings from CERs (with or without grafts), we implant sterile metal EEG recording electrodes with mounting screws (Plastics One) epidurally, one over the right fronto-parietal cortex for recording EEG from the cortex, and another over the left cerebellum as a reference electrode. To record EEG directly from the hippocampus, an intracranial stainless steel electrode (Teflon-coated except for the tip) with socket (Plastics One) is also placed into the right DG (Rao et al., 2006a). We also implant a couple of anchoring screws on the skull to secure the EEG electrodes with dental cement. The screws and electrodes are cemented in place, and electrode leads are attached to a microplug, which is then cemented to the animal’s head. The above implantation procedures are done in one surgery session (as detailed in Rao et al., 2006a). In grafting studies, this is typically done at ~2 to 3 months after the grafting.
Two weeks after the implantation surgery, each rat is placed in a Plexiglas cage, and the connector cable of the video-EEG system is fixed into the electrode pedestal on the rat’s head. The video-EEG system monitors simultaneously occurring behavior and EEG activity in awake, freely behaving rats with ad libitum access to food and water.
In our laboratory, the EEG recordings are done with a low-frequency filter (LF) set at 0.3 Hz, a high-frequency (HF) filter set at 35 Hz, and data rate set at 200 Hz. Furthermore, the SZAC detector component of AS40 is turned on to quantify the number of all high-amplitude spikes (first half-wave amplitude ratio set at ≥1.5 and 2nd half-wave amplitude ratio set at ≥2.5), and seizure events (high-frequency multispike complexes at 3 to 20 Hz and 50% faster than background), and/or high-voltage synchronized spike or wave activity (≥45% variation in amplitude, ≥36% variation in duration, amplitude ratio of 3, and lasting for ≥6 sec).
Both frequency and severity of EEG seizures with or without behavioral manifestation is continuously measured for a period of time (e.g. 2–6 weeks). An “EEG seizure with behavior” is a generalized electrographic seizure with accompanying motor seizure activity—e.g., unilateral or bilateral forelimb clonus, and rearing and falling (Stages III to V seizures). The various parameters for comparison across the age-matched animal groups (CERs receiving NSC or GPC grafts, CERs receiving sham-grafting surgery, CERs receiving neither grafts nor surgery and CERs receiving cyclosporine treatment alone) using one-way ANOVA include the following: (1) frequency of all SRS; (2) EEG-SRS with behavior; (3) EEG-SRS without behavior (electrographic SRS); (4) average duration of individual SRS; (5) total time spent in SRS; (6) number of high-amplitude spikes.
The EEG studies are labor intensive and clearly need personnel with expertise in appropriate implantation of electrodes and ability to connect the electrodes on the rat’s head to the video-EEG system through a connector cable without anesthetizing the animal. One pitfall of this procedure is that the number of rats initially implanted with electrodes may be reduced at the time of recordings due to malfunction of electrodes. Alternatively, electrodes may fall off due to accidents in the cage or during daily cyclosporine injections in rats receiving human GPCs. To circumvent these difficulties, it is important to include a larger cohort of CERs than required for statistics in each group. Furthermore, it will be necessary to individually house rats in cages with flat tops, which minimizes the loss/malfunction of electrodes due to accidents in the cage.
If CERs chosen for different groups (e.g. NSC or GPC graft groups) had undergone examination of cognitive function using one or two hippocampus-dependent cognitive tests (e.g. OLT) prior to the grafting or sham-grafting surgery (as described in an earlier section of this article), their pre-grafting/pre–sham grafting cognitive scores are available. In such scenario, the same tests can be employed at 2 to 4 months post-grafting or post-sham grafting surgery to ascertain changes in the cognitive or memory function with NSC/GPC grafting or sham-grafting surgery
Furthermore, regardless of pre-grafting behavioral tests, it will be important to perform a series of behavioral tests to examine cognitive, memory and mood function at 2–4 months post-grafting. In our laboratory, we employ the following behavioral tests for examining cognitive and mood function:
This test is performed in an open field box measuring 100 cm (L) × 100 cm (W) × 60 cm (H). Each animal is subjected to three successive trials in this test (Fig.6).
A day before the test, all animals need to be handled. Additionally, make sure that all chosen CERs explore the open field apparatus individually for 10–15 minutes. This pre-test exploration of the apparatus reduces anxiety in CERs on the day of testing. This was evidenced through increased time spent in exploration of objects by CERS in trials 2 and 3.
Export data such as times spent with novel and familiar objects and the total object exploration time from the software. Calculate percentages of the total object exploration time spent with the novel object vis-à-vis the familiar object. Compare these values within each animal group using two-tailed, unpaired Student’s t-test to determine the ability of animals for novel object recognition. The choice to explore the novel object over the familiar object reflects the ability of animal for recognition memory function. Typically, CERs (not receiving any therapeutic treatment) display a clear impairment in this recognition memory test, as they do not show affinity for the novel object in trial 3. Rather, they spend either nearly equal amounts of time with the familiar and novel objects or spend greater amount of time with the familiar object. If NSC or GPC grafting alleviates recognition memory dysfunction, then animals receiving NSC or GPC grafts would demonstrate ability for discriminating novel object from the familiar object by spending a greater amount of time with the novel object in trial 3.
This test is performed in an open field box measuring 100 cm (L) × 100 cm (W) × 60 cm (H). Each animal is subjected to three successive trials in this test.
A day before the test, all animals need to be handled. Additionally, make sure that all chosen CERs explore the open field apparatus individually for 10–15 minutes. This pre-test exploration of the apparatus reduces anxiety in CERs on the day of testing. This was evidenced through increased time spent in exploration of objects by CERS in trials 2–4.
Export data of trial 3 such as times spent with the object from trial 1 (i.e. novel object on pattern 2, NO on P2]) and the object from trial 2 (familiar object on pattern 2 [FO on P2] and the total object exploration time. Calculate percentages of the total object exploration time spent with the NO on P2 vis-à-vis FO on P2. Compare these values within each animal group using two-tailed, unpaired Student’s t-test to determine the ability of animals for pattern separation. Excellent pattern separation ability (i.e. ability to distinguish between similar experiences) in naïve rats is revealed through greater exploration of the object from trial 1 (i.e. NO on P2) than the object from trial 2 (i.e. FO on P2). Previous studies have shown that this task requires normal levels of dentate neurogenesis (Jain et al., 2012; Oomen et al., 2014; McAvoy et al., 2015). CERs (not receiving any therapeutic treatment) typically display no preference for the NO on P2, as they spend either nearly similar amounts of time with novel and familiar objects on P2 or spend greater amount of time with the familiar object, implying loss of ability for pattern separation, which is not surprising as chronic epilepsy is associated with a great decline in dentate neurogenesis (Hattiangady et al., 2004, 2010; Kuruba et al., 2009). If NSC or GPC grafting alleviates pattern separation dysfunction, then animals receiving NSC or GPC grafts would demonstrate ability for discriminating NO on P2 from FO on P2 by spending a greater amount of time with the NO in trial 4.
Following the completion of all behavioral measures described above, animals are perfused with 4% paraformaldehyde for analyses of the yield and differentiation of NSC or GPC graft derived cells. This comprises tissue fixation, tissue processing for cryostat sectioning, immunostaining and/or immunofluorescence procedures, quantification of the yield of graft-derived cells using stereology, and measurement of the percentages of the graft-derived cells that differentiate into NeuN+ neurons, GABA-ergic interneurons, and subclasses of GABA-ergic neurons expressing various markers, astrocytes, astrocytes expressing neurotrophic factors such as the GDNF, oligodendrocyte progenitors, and mature oligodendrocytes.
For more detail on the techniques described below, see Hofman (2002).
For more detail on the techniques described below, see Rao and Shetty (2004).
Select every 10th or 15th section through the entire hippocampus for analyses of the yield of human GPC graft-derived cells. Process these serial sections for human nuclear antigen (HNA) immunostaining in 24-well plates.
For quantifying the yield of graft-derived cells, we typically count all BrdU+ or HNA+ cells in every 10th or 15th section through the entire anterior-posterior extent of the hippocampus using the optical-fractionator counting method in the StereoInvestigator system (MicroBrightField Inc.; http://www.mbfbioscience.com/). The overall yield of graft-derived cells may be expressed as the total number of engrafted cells per hippocampus or as the percentage of the total injected cells per hippocampus (i.e., the total number of BrdU+/HNA+ cells placed at four sites in each hippocampus).
It is important to note that the yield of graft-derived cells does not represent the absolute survival of cells that are originally implanted. However, assessment of the yield does give an idea about the numbers of graft-derived cells that engraft for prolonged periods when a given number of NSCs or GPCs are implanted. Gauging the yield is important for NSC and GPC grafts, as quantification of the absolute graft cell survival is difficult due to the likelihood that both cell proliferation and cell death occur in these grafts. Thus, one can conclude that the overall engraftment is robust if the yield of graft-derived cells is close to or greater than the number of cells originally implanted. A yield greater than the number of cells initially injected can happen if the NSCs or GPCs divide a couple of times after the grafting. The major aspects of the optical fractionator cell counting methodology are described below.
The StereoInvestigator system consists of a color digital video camera (Optronics Inc., http://www.optronics.com/) interfaced with a Nikon E600 microscope. In each hippocampus, the BrdU+ cells are counted from randomly and systematically selected frames (e.g., each measuring 150 × 150 μm, 0.0225 mm2 area) in every 10th section using the 100× oil-immersion objective lens. The numbers and densities of frames are determined by entering the parameter grid size (e.g., 60 × 60 μm) in the optical fractionator component of the StereoInvestigator system (Rao and Shetty, 2004). Although 30-μm-thick sections through the hippocampus are cut using a cryostat, the section thickness at the time of counting reduces to 40% to 50% of the initial thickness because of the shrinkage of the tissue with the BrdU immunostaining. Hence, at the time of data collection, the thickness of sections varies from 12 to 15 μm.
It is very important to check the thickness of each section used for counting before entering the section thickness, as shrinkage can vary between sections.
For commencing cell counting in the each chosen section, open the serial section manager dialog box and enter different parameters such as the number of sections, section evaluation interval, section cut thickness, mounted section thickness, and the starting z level.
Mark the contour of the transplant area (i.e., the area of the hippocampus containing the graft-derived [i.e., BrdU+ or HNA+] cells) using the 4× or 10× objective lens and the tracing function of the StereoInvestigator.
Activate the optical fractionator component of the program and select the numbers and locations of the counting frames (using a systematic random sampling scheme) and the counting depth by entering parameters such as the grid size (e.g., 60 × 60 μm), the thickness of top guard zone (e.g., 4 μm), and the optical dissector height (e.g., 6 μm).
The number of frames per section counted varies, as the overall area comprising the graft-derived cells varies from section to section. This sampling scheme is consistent with the principle of the optical fractionator counting method in that the sample concentration must remain constant for each section, implying that the number of unit volumes to be counted per volume of the structure on a given section needs to be a constant ratio from section to section. Thus, counting of cells from randomly and systematically chosen frames in every 10th or 15th section through the hippocampus guarantees that effectively every BrdU+/HNA+ cell within the transplant area has equal odds of being counted. This is imperative because the scattering of BrdU+/HNA+ cells within the graft area is visibly heterogeneous.
A computer-driven motorized stage then allows the section to be analyzed at each of the counting-frame locations. In every counting-frame location, identify the top of the section, and then move the plane of the focus to a depth of 4 μm (guard zone) to get rid of the problem of uneven section surface. This plane serves as the first point of the counting process. Count all BrdU+/HNA+ cells that come into focus in the next 6-μm section thickness if they are entirely within the counting frame or touching the upper or right side of the counting frame.
Repeat the above protocol for every chosen section in each hippocampus. Following collection of data from all sections, select “display probe run list” to obtain the total number of graft-derived (BrdU+/HNA+) cells. The extent of migration of graft-derived cells can be assessed indirectly by obtaining the total volume of the hippocampal tissue comprising the graft-derived cells from the “display probe run list.” Based on the different parameters that are entered at the beginning of the counting and the number of cells sampled at different counting locations in all sections, the StereoInvestigator program calculates the total number of BrdU+ cells per hippocampus by utilizing the optical fractionator formula, N = 1/ssf.1/asf.1/hsf. EQ− (Dorph-Petersen et al., 2001). The abbreviation “ssf” stands for the section sampling fraction, which is 10 in this example; “asf” symbolizes the area sampling fraction, which is calculated by dividing the total areas sampled by the total area of the transplant (i.e., the sum of transplant areas in every 10th section); “hsf” stands for the height sampling fraction, which is calculated by dividing the height sampled (which is 6 μm in this example) by the section thickness at the time of analysis (i.e., 12 to 15 μm); “EQ− denotes the total count of particles (i.e., BrdU+ or HNA+ cells) sampled for the entire hippocampus. Figure 10 illustrates the distribution of cells derived from NSC grafts along the septo-temporal axis of the hippocampus of a chronically epileptic rat (Waldau et al., 2010).
Following collection of data (such as the total number of BrdU+/HNA+ cells per hippocampus), check the Gundersen coefficient of error (CE) for each sample. The CE values that are <0.05 suggest that the counts obtained are valid.
If larger CE values are seen for some samples, repeat the counting process by modifying the counting parameters, which may include increasing the number of sections (e.g., every 5th section instead of the every 10th section), altering the grid size to increase the number of sites per section, and changing the counting-frame dimensions to increase the probability of counting more cells at each counting location.
Cells derived from the NSCs are typically heterogeneous, and each type of cell derived from NSCs has a unique function. Therefore, it is vital to quantify the fraction of graft-derived cells that become mature neurons (e.g., NeuN+ cells), specialized neurons (e.g., GABA-ergic neurons), mature astrocytes (e.g., S-100β+ cells), oligodendrocyte progenitors (e.g., NG2+ cells), and mature oligodendrocytes (e.g., O1+ and Rip+ cells). In addition, as NSC differentiation might depend on factors in the host tissue or the location of engraftment, it is also important to assess the fraction of graft-derived cells that persist as NSCs (e.g., Sox2+ or nestin+ cells) in the host brain. This will provide clues as to whether cells derived from NSC grafts facilitate repair and regeneration of the host brain area via replacement of lost cells or through mechanisms such as the release of neurotrophic or other beneficial factors.
Cells derived from the human GPCs are expected to be mostly homogeneous, as a vast majority of cells derived from GPCs are expected to differentiate into GABA-ergic interneurons. However, smaller fractions of cells from GPCs may also differentiate into astrocytes, oligodendrocytes or may remain as progenitors expressing NKX2.1. Therefore, it is vital to quantify the fraction of graft-derived cells that become mature neurons (e.g., NeuN+ cells), interneurons (e.g., GABA-ergic neurons), subclasses of interneurons (NPY, SOM, PV, CR, CBN, reelin etc.), mature astrocytes (e.g., S-100β+ cells), oligodendrocyte progenitors (e.g., NG2+ cells), and mature oligodendrocytes (e.g., O1+ and Rip+ cells) and GPCs (NKX2.1). In addition, to rule out the continued proliferation of NKX2.1+ GPCs in the host brain, it will be necessary to quantify fractions of graft-derived cells expressing markers of proliferation (e.g. Ki-67). Furthermore, as GPCs derived from hiPSCs are used for grafting, possible occurrence of pluripotent stem cells need to be investigated among graft-derived cells using appropriate markers (e.g. Oct-4).
The differentiation of graft-derived cells is typically assessed through a standard dual immunofluorescence method and z section analyses in a confocal microscope, by identifying both the graft cell marker (e.g. BrdU or HNA) and the chosen neural cell marker (such as NeuN, GABA, S-100β+, NG2, O1, Rip, Sox2, nestin, NPY, SOM, PV, CR, CBN, reelin, NKX2.1, Ki-67, Oct-4 etc.).
As the dual-immunofluorescence approach can be done using different protocols, we restrict our description mainly to the listing of primary and secondary antibody combinations that we typically use in our studies (see Table for details). We employ free-floating sections in 24-well tissue culture plates and sequentially visualize the two antigens. The first antigen is typically the graft cell marker (i.e., BrdU or HNA), whereas the second antigen is the chosen neural cell/NSC marker. The overall procedure takes ~3 days to complete for each combination.
Following dual-immunofluorescence staining, we perform z section analyses in a confocal microscope within a couple of weeks to prevent the fading of fluorescence over time. We obtain z sections at 1-μm intervals and perform orthogonal analysis to confirm the presence of dual antigens in all graft-derived cells that are selected for quantification. This will facilitate the assessment of the percentages of graft-derived cells that express different neural antigens. Examples of the differentiation of cells derived from MGE-NSC grafts into different neural phenotypes in the chronically epileptic hippocampus are illustrated in Figure 11.
In addition to contributing new neurons synthesizing the inhibitory neurotransmitter GABA and new astrocytes synthesizing the anticonvulsant protein GDNF (Waldau et al., 2010), NSC of GPC grafting into the epileptic hippocampus might also influence the host cells and neurons, as observed in other disease models. For example, NSCs placed into the substantia nigra and caudate nuclei of monkeys exhibiting Parkinson’s disease can induce behavioral recovery with differentiation of only a small fraction of grafted NSCs into dopaminergic neurons (Redmond et al., 2007). As large numbers of grafted NSCs became differentiated into astrocytes and expressed multiple neurotrophic factors including GDNF, it appeared that NSC grafting–mediated changes to the host milieu enhanced the function of the existing dopaminergic neurons (Redmond et al., 2007; Sanberg, 2007). Considering this, it is important to examine the effects of NSC and GPC grafts on the host microenvironment and host cells, in addition to examining the differentiation of graft-derived cells into desired neuronal or glial phenotypes. In our previous study, we utilized a standard dual-immunofluorescence approach to assess the changes in the function of S-100β+ host astrocytes with NSC grafting into the hippocampus of CERs (Waldau et al., 2010). The results showed that NSC grafting into the hippocampus of CERs considerably restored GDNF expression in the S-100β+ host hippocampal astrocytes (Waldau et al., 2010). However, it is likely helpful to use high-throughput genomic (such as microarray and qRT-PCR arrays) and proteomic approaches in the future to understand the effects of NSC or GPC grafting on multiple pathways in the host hippocampus of CERs. The methods for these approaches are not described here, as they are beyond the scope of this article.
For culture recipes and steps, use sterile tissue culture–grade water. For other purposes, use deionized, distilled water or equivalent in recipes and protocol steps. For suppliers, see SUPPLIERS APPENDIX.
In our laboratory, we use an anesthetic cocktail comprising ketamine (50 mg/ml), xylazine (4.5 mg/ml), and acepromazine (0.4 mg/ml) at a dose of 0.7 ml/kg body weight.
The cocktail may be stored up to 2 weeks at 4°C.
Add 6.18 g boric acid (Sigma) to 1000 ml of distilled water and adjust the pH to 8.5.
For 50 ml:
For 50 ml:
Add 12.11 g Trizma base (Sigma) and 8.77 g of sodium chloride (Sigma) to 100 ml of distilled water and adjust the pH to 7.5.
Epilepsy, typified by spontaneous recurrent seizures (SRS) due to hyperexcitability and synchronization of activity within populations of neurons, affects >50 million people (Strine et al., 2005). Temporal lobe epilepsy (TLE), characterized by progressive development of complex partial seizures and hippocampal neurodegeneration, is seen in >30% of epileptic patients (Manford et al., 1992). While the etiology of TLE is unknown in most cases (McNamara, 1999), it is typically seen after an initial precipitating injury such as status epilepticus (SE), brain injury, tumors, meningitis, encephalitis, and febrile seizures (French et al., 1993; Mathern et al., 1995; Lewis, 2005). Seizures in TLE originate from temporal-lobe foci, which are associated with learning and memory impairments, reduced dentate neurogenesis, and depression (Devinsky, 2004; Hattiangady et al., 2004; Detour et al., 2005; Pirttilä et al., 2005; Hattiangady and Shetty, 2010; Hattiangady et al., 2011). Nearly 35% of patients with TLE exhibit seizures that cannot be controlled by antiepileptic drugs (Litt et al., 2001), and memory difficulties are a frequent cognitive complaint in patients with chronic epilepsy (Vannest et al., 2008). Although surgical resection of the epileptic hippocampus gives better seizure control, this option is often associated with significant cognitive impairments (Helmstaedter et al., 2008). Hence, there is a pressing need to develop alternative therapeutic approaches that greatly diminish both frequency and intensity of SRS in patients with chronic TLE on a long-term basis.
Pharmacological SE animal models using KA or pilocarpine are popular for studying TLE, as they replicate several features of TLE. Kainic acid (KA) is a widely used excitotoxin in studies of the hippocampus (Nadler et al., 1978; Sperk et al., 1994; Shetty and Turner, 1996, 1999; Rao et al., 2006a,b, 2007; Rao et al., 2008; Shetty et al., 2012). Intraperitoneal (i.p.) administration of KA in rat induces degeneration of dentate hilar neurons and fractions of CA1 and CA3 pyramidal neurons, deafferentation of dentate granule cells due to loss of hilar mossy cells, and reduction in number of GABA-ergic interneurons in the hippocampus. These changes are followed by aberrant sprouting of dentate mossy fibers into the dentate supragranular layer (DSGL), hyperexcitability in the hippocampus, and a chronic epileptic condition, typified by SRS, learning and memory deficits, and depression (Ben-Ari, 1985; Letty et al., 1995; Hellier et al., 1998, 1999; Buckmaster and Dudek, 1999; Wuarin and Dudek, 2001; Rao et al., 2006a). Since the above changes in the adult hippocampus closely resemble human mesial TLE, the i.p. KA model has been widely used for studying TLE. This model has also been found to be useful for studying the effects of cell grafts on chronic seizures (Rao et al., 2007; Waldau et al., 2010). Thus, the intraperitoneal KA model is suitable for testing the efficacy of NSC grafts in alleviating chronic epilepsy characterized by spontaneous seizures and learning and memory impairments.
It is believed that an increased excitatory neurotransmission found in the epileptic hippocampus is partly due to a reduced number of GABA-ergic interneurons, loss of functional inhibition, and diminished numbers of GABA-ergic terminals (Ribak et al., 1986; Cornish and Wheal, 1989; During et al., 1995; Shetty and Turner, 2000, 2001; Shetty et al., 2009). From this perspective, the idea of restraining SRS in the epileptic hippocampus via grafting of cells that just release the inhibitory neurotransmitter GABA at the seizure focus has received considerable attention (Löscher et al., 2008). For instance, grafting of GABA-soaked beads, immortalized GABA-ergic cells, cells that are engineered to produce GABA, and fetal GABA-ergic cells derived from medial and lateral ganglionic eminences and GABA-ergic progenitors derived from hESCs into the epileptic foci have been shown to reduce seizures in a variety of animal models (Löscher et al., 1998; Gernert et al., 2002; Thompson, 2005; Castillo et al., 2006; Hattiangady et al., 2008; Hunt et al., 2013, Henderson et al., 2014; Cunningham et al., 2014). Thus, grafting of GABA-producing cells into the epileptic brain has considerable promise for restraining seizures. However, a routine clinical application of human fetal cells for TLE may not be feasible because of the difficulty in obtaining the required quantity of human fetal cells, and ethical issues (Turner and Shetty, 2003). With reference to the other cell types, it is unknown whether immortalized GABA-producing cells have the ability for long-term survival in the chronically epileptic brain, as they are expected to perform as GABA pumps rather than undergo synaptic integration with the host neurons (Löscher et al., 2008). Therefore, there is a need to find types of cells that are capable of providing an unlimited source of donor cells for grafting and have the ability to give rise to large numbers of GABA-ergic interneurons that survive for prolonged periods after grafting into the epileptic brain.
From the above perspective, it appears that NSCs expanded from the embryonic MGE (Waldau et al., 2010) and the postnatal and adult subventricular zone (SVZ) are attractive candidates as donor cells for grafting therapy in TLE. This is because these cells can (i) be expanded for prolonged periods in culture without losing their multipotential property; (ii) give rise substantial numbers of GABA-ergic interneurons; and (iii) release a multitude of neurotrophic factors (Shetty and Hattiangady, 2007; Waldau et al., 2010). Additionally, MGE-like GPCs obtained from sources such as hESCs and iPSCs may also be used once their potential to give rise to full-fledged GABA-ergic neurons and/or to release multiple neurotrophic factors is validated. However, prior to the clinical application of NSCs or GPCs obtained from any source for TLE, it will be necessary to rigorously analyze the ability of the chosen NSCs and GPCs to restrain SRS on a long-term basis using both behavioral and EEG analyses in animal models of TLE. In addition, the efficacy of different NSC or GPC grafts for improving cognitive and mood function in TLE need to be examined. Particularly, detailed anatomical, electrophysiological, and molecular biological analyses of the efficacy of grafting of NSCs and GPCs for easing SRS and cognitive and mood dysfunction in chronic TLE on a long-term basis are needed before considering the clinical application of NSC or GPC therapy for TLE.
Generating CERs via SE induction is a laborious process that needs patience and close attention to every step. Some of the issues and precautions to be taken are described here. (1) The sensitivity to KA may vary among F344 rats purchased from different sources. Hence, it is ideal to use rats obtained from the same source for all experiments in a particular study. (2) It is important to use KA from a single source for the entire study, as KA from different vendors seems to vary in terms of potency to induce acute seizures or SE. It is recommended that the potency of KA from a particular source be checked in a pilot experiment and that the appropriate dose for inducing SE prior to initiating SE studies in a large number of animals be determined. (3) The use of cages having flat tops fitted with foam lining on the inner side is useful for preventing injury to rats in the event of bouncing seizures after the onset of SE. (4) If a rat is having very intense seizures such as bouncing seizures after the onset of SE (which can happen suddenly), carefully handling the rat and inducing hypothermia (e.g., by placing the rat in an empty cage that was kept on a larger container filled with ice) for a few minutes stops such intense seizures and prevents mortality. (5) Injecting diazepam (5 mg/kg body weight) after 2 or 3 hrs of seizure activity considerably reduces the mortality of rats and does not interfere with the occurrences of SRS in the chronic phase after SE. Without diazepam administration, the overnight mortality after SE will be significantly higher, though the rats that survive might exhibit robust chronic epilepsy (characterized by greater frequency and intensity of SRS). On the other hand, if diazepam is administered very early after the onset of SE (e.g., immediately after the first stage V seizure or before the occurrence of stage V seizure), only a small fraction of rats will develop chronic epilepsy at 4 to 6 months post-SE.
Appropriate animal care after SE is another requirement for reducing SE-related mortality. After SE, animals appear weak, lethargic, and anorexic, and do not seem to eat hard pellets or drink from the water bottle. Therefore, preventing dehydration via subcutaneous administration of fluids (such as saline or Ringer’s lactate solution) and providing soft rat chow or transgel within the cage in a dish clearly helps in minimizing mortality in the recovery phase after SE.
The various methods such as dissection of MGE tissues from the embryonic brain, expansion of NSCs in culture, expansion of NKX2.1+ GPCs from hiPSCs, and preparation of NSC and GPC suspension for grafting can be performed only by personnel with specific expertise in these aspects. Trituration of neurospheres is a highly skilled technique, as too-gentle trituration does not dissociate the neurospheres well and too-harsh trituration will kill a large number of cells. Triturating neurospheres in a proliferation medium minimizes cell death, in comparison to triturating in any other medium. It is important to select the NSC or GPC suspension with a good viability index (e.g., >80%) for grafting studies because excellent viability of donor cells in the suspension at the time of grafting appears to be an important prerequisite for successful engraftment into the host brain and for obtaining higher yields of graft-derived cells at extended time-points after grafting. A cell suspension with a poor overall viability (e.g., <70%) should not be used for grafting, as greater numbers of dead cells in the graft might initiate an inflammatory reaction in the host brain. Higher-volume injections at a single site in the hippocampus should be avoided, as this typically causes tissue damage, opening of the tissue cleavage planes (e.g., the hippocampal fissure), and flow of the injected fluid mostly into the lateral ventricle. Based on our experience, injection of a 1-μl volume comprising 80,000 to 100,000 live cells (per site) into the hippocampus results in good engraftment of graft-derived cells and does not seem to cause tissue damage at the grafted site. Therefore, preparation of NSC or GPC suspension with a high density of cells is required, which clearly requires expansion of NSCs and GPCs in a large number of flasks.
Typically, CERs do not exhibit SRS while handling. However, occasionally, it is possible that a CER can have a seizure in the middle of a behavioral test. In such cases, it is necessary to remove the rat quickly from the testing apparatus to prevent any injury. Such rats should be observed closely for a while and new trials should not be given in the immediate post-SRS confusion period. However, such rats may be re-tested after a delay of 2–3 days, depending upon the type of test. Furthermore, some CERs might show a tendency to drown in pre-trials of FST. Such rats should be excluded from FST.
When all guidelines and protocols described in this article are faithfully followed, success of the experiment (i.e., testing the efficacy of NSC of GPC grafting for restraining SRS and improving cognitive and mood function in chronic TLE) in terms of drawing reliable conclusions is ensured. Additionally, the experimental results can be reproduced with sufficiently stringent adherence to the guidelines. However, errors such as inclusion of rats from different sources, the use of KA from different sources, an inappropriate selection of CERs (e.g., a group of rats with widely differing frequency and intensity of SRS) for grafting, grafting donor NSC or GPC suspension with a poor viability of cells, larger-volume cell suspension injections into the hippocampus, and withdrawing the needle immediately after the injection of the cell suspension introduce significant confounds, and the results obtained from such experiments will be unreliable for meaningful conclusions, and non-reproducible.
The experiment described here for testing the efficacy of NSC or GPC grafting for restraining SRS and improving cognitive function using an animal model of chronic TLE is clearly a long-term experiment. In our experience, animal aspects of the work such as induction of SE, characterization of SRS, and cognitive dysfunction in CERs, as well as expansion and grafting of NSCs and GPCs, post-grafting analyses of SRS (via direct observations/video recordings and video-EEG recordings), and cognitive and mood function (via OLT, NORT, PST, SPT, ERDT and FST), take ~1 year to complete. Following this, processing of tissues via perfusion, serial section cutting of brain tissues, histology, immunostaining, stereological counting of graft-derived cells, dual immunofluorescence and confocal microscopic analyses of the phenotype of graft derived cells, and analyzing the effects of grafts on host cells and microenvironment would need significant additional amounts of time. The overall time required for histological processing, immunostaining, and quantification depends on the number of animals in different groups, numbers of experimental groups, and the type of analysis undertaken. Stereotaxic surgery can be laborious, as injections are made into multiple sites in the hippocampus. However, with practice, transplantation can be performed on six animals in one surgery session if two stereotactic devices are available. Measuring SRS either via direct observation or from the recorded video requires strong observation skills and an ability to distinguish SRS from the normal grooming behavior of rats throughout the observation period. Furthermore, enthusiastic research personnel/scientists who are committed to this type of long-running and labor-intensive (but very interesting and clinically relevant) experiment are needed for successfully completing these studies.
This work was supported by grants from the Department of Defense (PRMRP award, W81XWH-14-1-0558 to A.K.S.), the Department of Veterans Affairs (VA Merit Award, I01 BX002351 to A.K.S.) and the State of Texas (Emerging Technology Funds to A.K.S.). The authors also acknowledge the previous funding from the National Institute of Neurological Disorders and Stroke (R01 NS054780, R01 NS043507, and R01 NS36742 to A.K.S.) and the Department of Veterans Affairs (Merit Review Awards to A.K.S.) for studies on animal models of temporal lobe epilepsy in SHETTY LAB. The authors wish to thank Vandana Zaman, Muddanna Rao, Ben Waldau, Ramkumar Kuruba, and Bing Shuai for their outstanding contributions to original articles on neural cell grafting in animal models of temporal lobe epilepsy published from SHETTY LAB, and Mr. Luciano Silveira for his contribution to preparation of figures. Dr. A. K. Shetty is also Research Career Scientist at the Olin E. Teague Veterans’ Medical Center, Central Texas Veterans Health Care System Temple, Texas. Dr. Upadhya is an Assistant Professor on study leave from S.D.M. College of Ayurveda, Udupi, India.