Chemicals
Anaesthetics: Isoflurane (Pharmachem Queensland); Urethane >99% (Sigma-Aldrich, Australia, CAS number 51-79-6)
Heparin: Heparin Sodium (Porcine Mucous) Injection BP (HameIn pharmaceuticals GmbH, Germany)
Lithium Chloride Anhydrous LR (Chem-Supply, CAS number: 7447-41-8)
Optimal Cutting Temperature media: TissueTek OCT from Sakura Finetek
Isotonic sodium chloride: 0.9g/100ml
Sodium Chloride (Baxter Healthcare, New South Wales)
Digoxin 98.9% purity (Sigma-Aldrich, CAS number: 20830-75-5)
Animals and surgical procedures
Sprague Dawley rats were sourced from The Biomedical Sciences Animal Facility at The University of Melbourne. All rats were subjected to a 12h light/ dark cycle with ad libitum access to food and water. All experimental procedures were conducted in accordance with National Health and Medical Research Committee (NHMRC) guidelines, the ARRIVE guidelines and approved by The University of Melbourne Ethics Committee, Ethics ID 1914793 (Entry of Anti-Epileptic and Psychotic Drugs into the Developing Brain).
As present experiments required administration of lithium by intraperitoneal (i.p.) injection, the more neutral lithium chloride salt rather than the strongly alkaline clinical preparation of lithium carbonate was used (see Introduction).
Animals were randomly allocated to two experimental protocols:
- acute experiments, where animals received a single injection of lithium chloride (LiCl),
- long-term exposure experiments where animals received multiple injections of LiCl over several days.
Therapeutic concentrations of lithium in the blood of patients on lithium therapy are in the range of
0.4 -1.2mmol/l (6), expressed in the present study as mM. Doses and injection volumes were standardised to the animal’s body weight (3.2mg/kg lithium, see Figure 1A) to achieve the lowest therapeutic concentration of lithium in plasma and injection volumes were limited to <1 % of body weight to avoid any significant increase in the circulating blood volume. LiCl was dissolved in 0.9% sterile sodium chloride prior to intraperitoneal (i.p.) injection.
Ages, weights and numbers of animals used of A. untreated, B. acute, C. long-term treated fetuses/ pups and D. Dams. E= embryonic, P= postnatal, n, number of animals at each age, SD=standard deviation. For sex and weight of individual animals see Supplementary file Tables 1-2. Numbers in brackets indicate the number of litters used for each age group. Note that for E18, the pup’s size is given in mm of Crown-Rump (CR) lengths as this is a more accurate parameter than weight for fetal staging.
Untreated animals were litter-mates of treated animals (acute experiments) or were obtained from other studies, including untreated control adult females (37, 38). All individual lithium measurements are shown in Supplementary file 1, Tables 3-5.
Injection protocols
- Fetal animals
Fetal animals were exposed to lithium via placental transfer only. This was achieved by i.p. injections into time mated pregnant dams. For acute exposure experiments, E18 pregnant rats were administered a single i.p. injection of 3.2mg/kg lithium. For long-term exposure experiments, twice daily 3.2mg/kg lithium i.p. injections (early morning and late afternoon) were administered to dams from E15-E18.
(ii) Postnatal animals
For acute experiments, individual pups and dams were injected i.p. with 3.2mg/kg lithium and left for 90-120min (duration where plasma and CSF levels are stable, see Figure 1B-D). For long-term experiments, pups were exposed to lithium only via breast milk from dams injected twice daily (early morning and late afternoon) with a standard dose of lithium (3.2 mg/kg body weight) and collected after 2, 4, 7, 12 or 16/17 days of exposure to breast milk. Dams were collected together with P16/17 pups (17 or 18 days of treatment respectively). Pups in the long-term experiments were separated from the dam 2-3h after dam’s last injection.
Untreated control animals were age-matched to all experimental animals.
Treatment protocols and surgery
Fetal animals
Pregnant rats were deeply anaesthetised (urethane 2.5g/kg i.p.), a tracheal cannula inserted to maintain the airway and left femoral artery catheterised for collection of maternal blood samples. Animals were placed on heated plate (33ºC) and a uterine horn exteriorised via a mid-line abdominal incision. Pups, starting at the ovary end of the uterine horn, were removed from their amniotic sac and samples collected as described below. Any pups with visible signs of haemorrhages were not collected, but their position in the uterine horn was recorded.
Postnatal animals
Pups were terminally anaesthetised with an overdose of inhaled anaesthetic (>5% isoflurane), before samples were collected (see below).
Inhibition of sodium/ potassium ATPase in P4 pups
Pups at P4 were used for the experiment investigating if partial blocking of Na+/K+ ATPase could influence lithium transfer into the CSF. Seven pups from two separate litters (3 and 4 respectively) were injected with digoxin (300mg/kg body weight in 75µl, dissolved in ethanol followed by isotonic sodium chloride solution) and 7 matched littermates were injected with equal volume of isotonic sodium chloride solution, all injections were given i.p. Pups were carefully monitored for 30min, before they were given the same lithium injection protocol as other postnatal animals (see injection protocol). Samples were all collected within 90-105min after lithium injection and all samples were coded for blinded determination of lithium analysis (see below).
Collection and processing of blood and CSF samples
Blood samples were taken directly from the right ventricle of the heart into heparinised glass capillaries. For the acute experiments in pregnant rats, a small sample of maternal arterial blood (0.2ml) was collected from the femoral arterial catheter at the same time as each pup was removed and the catheter then flushed with an equal volume of isotonic sodium chloride solution to maintain circulating volume (~2ml of blood was collected from each dam in total).
Blood samples were centrifuged for 5min (2,000xg) and plasma separated. Samples of CSF were collected from the cisterna magna using a glass micropipette, with care taken to avoid rupturing blood vessels. Any sample that was visibly contaminated with blood was discarded (contamination as small as 0.2% can be visibly detected, (39). Plasma and CSF samples were quantified using two different methods: inductively coupled plasma-mass spectrometry (ICP-MS) and laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS).
ICP-MS
Frozen aliquots of undiluted plasma and CSF were thawed at room temperature. Plasma (100μl in adults and 10-20μl for E18 and P4 animals) and 10µl of CSF were aliquoted into separate 1.7ml microcentrifuge tubes. Plasma and CSF were diluted in 3.5% (v/v) nitric acid (HNO3, 70% Analytical grade from Ajax Finechem) in Milli-Q H2O (18.2 MΩ; Merk Millipore, Australia) (1:10 and 1:100 respectively to a final volume of 1ml) and digested by heating to 90°C for 10min. Samples were briefly vortexed and centrifuged at 20,000xg for 20min and supernatant was immediately transferred to a new 1.7ml microcentrifuge tubes. Blanks were prepared in an identical manner and 1ml of 3.5% HNO3 blanks were also aliquoted for analysis.
LA-ICP-MS
Plasma samples (0.5μl) diluted in 0.9% NaCl, 1:10 v/v and undiluted CSF were spotted in onto a glass slide in triplicate (Superfrost glass slides, Thermo Scientific) and air dried for 24h before analysis.
Collection and processing of brain samples
Whole brains, including olfactory lobes, were dissected out of the skull and bisected along the sagittal plane into two equal halves. Each half was randomly assigned to either extraction and quantification of lithium (using ICP-MS or LA-ICP-MS) or sectioning (for LA-ICP-MS).
2. Lithium extraction for ICP-MS
Approximately 50mg of tissue was transferred to a 1.7ml microcentrifuge tube and lyophilised. Both the wet weight and dry weight of the tissue were measured to report elemental concentration in terms of g of wet and dry tissue weight. Lyophilised tissue was digested in 50µl of 70% (v/v) HNO3 (70% Analytical grade from Ajax Finechem) at 90°C for 20min. Samples were allowed to cool prior to addition of 50µl of 30% (v/v) hydrogen peroxide (H2O2) (30% Hydrogen Peroxide, Analytical grade from Merck) and heated to 70°C for 15min. Samples were allowed to cool then diluted to a final volume of 1ml with 1% (v/v) HNO3 in Milli-Q H2O. Samples were briefly vortexed and centrifuged at 20,000xg for 25min and the soluble material was immediately transferred to a new 1.7ml microcentrifuge tube. Preparation blanks were prepared in an identical manner. Water blanks were also taken during the measurement.
Inductively coupled plasma-mass spectrometry (ICP-MS)
An Agilent 8900 triple quadrupole ICP-MS (Agilent Technologies) was tuned and optimised using a tuning solution containing 1 μg/l of cerium (Ce), cobalt (Co), lithium (Li), thallium (Tl) and Y in 2% (v/v) HNO3 (Agilent Technologies, Australia). The ion intensity at m/z 7 (lithium) was monitored in a no collision gas analysis mode. The instrument was calibrated using a 10-point calibration curve for lithium using commercially available multi-element standards at 0, 1, 5, 10,25, 50, 100, 250, 500 and 1000 parts per billion (ppb) in 1% HNO3 (Multi-Element Calibration Standard 2A, Agilent Technologies, USA). Yittrium (89Y) (Agilent Technologies) was used as an internal reference elemental standard at a concentration of 0.1μg/ml and used to normalise recovery across all samples. All samples, calibration standards and internal standards were introduced to the nebuliser using a peristaltic pump and T-piece for sample mixing at the flow rate of 0.4 ml/min. The ICP-MS operating parameters were established according to the manufacturer’s guidelines and other parameters were optimised for lithium in a batch specific mode prior to each experiment are as follows:
ICP-MS operating parameters
|
|
Scan type
|
Single Quad
|
RF Power
|
1550 W
|
RF Matching
|
1.2 V
|
Nebulizer Gas
|
1.09 L/min
|
Extract 1
|
-12 V
|
Extract 2
|
-215 V
|
Omega Bias
|
-95 V
|
Omega Lens
|
5.2 V
|
Deflect
|
10.8 V
|
Collision Gas
|
No Gas
|
Oct P bias
|
-8 V
|
Limits of detection (LOD: 0.013μg/l) and quantification (LOQ: 0.083μg/l) calculated as 1x SD of 10 replicate blanks were obtained. Seronorm™ Trade Elements Serum L-1 and L-2 (Sero, Norway) were used to externally assess analytical performance and sample recovery was within ± 20% RSD of manufacturer’s expected values.
3. Lithium extraction for LA-ICP-MS
Half of the brains designed for the extraction were frozen on dry ice and stored at -80 until use. Brain samples were defrosted for a minimum of 30min at room temperature, homogenised in isotonic sodium chloride solution (1:2 w/v) using a glass pestle, mixed vigorously and incubated at 37 for 20min. Samples were mixed briefly, centrifuged for 10min (2,000xg) and supernatant collected (total extracted volumes were recorded). The supernatants were diluted with isotonic sodium chloride (1:5 and 1:10 v/v) and replicates spotted onto glass slides, air dried and analysed using LA-ICP-MS (see below).
Laser ablation-inductively coupled-mass spectrometry (LA-ICP-MS)
Sectioning of brain for LA-ICP-MS: the frozen half brain samples (see above) were immediately placed in a mould containing Optimal Cutting Temperature media (OCT), frozen on dry ice and stored at -80 until sectioning. Frozen brain blocks were temperature-equilibrated for at least an hour before sectioning (30µm serial sagittal sections) in a cryostat. Sections were mounted on glass slides and air dried overnight before LA-ICP-MS analysis
Samples (0.5 l “droplets” or 30μm brain sections) were air dried on standard microscope slides for 24h before being placed into a 10×10 cm ablation cell where the laser was tuned to the topography and dimensions of the sample slides together with matrix-matched elemental standards for quantitative analysis (40). Brain sections were ablated with a 213nm laser (NWR213 ablation system, Kennelec Scientific) by a series of rasters using a 60 x 60μm square spot size and a scanning speed of 240μm/s for low resolution scans, and 30 x 30μm square spot size and a scanning speed of 120μm/s for higher resolution scans. For droplet (CSF, plasma and brain homogenate) analysis 100μm2 spot size and speed of 200μm/s was used. Ablated material was swept into Agilent 8800 QQQ-ICP-MS (Mulgrave, Victoria, Australia) by argon gas flow at 1.2l/min and directed through the plasma torch for ionisation. Ionised material was analysed for lithium (7Li), carbon (13C), sodium (23Na), phosphorus (31P), potassium (39K) and iron (56Fe) for a dwell time of 0.045 s per element. Carbon and phosphorus were recorded for tissue structure identification and iron for detection of residual blood in the tissue and CSF. Approximate analysis time for each brain section was 5-6h and for each microscope slide of droplets (100-150 droplets on average) 20-24h. Two dimensional elemental maps were constructed using the Iolite analysis software (School of Earth Sciences, University of Melbourne) operating under the Igor Pro suite (41).
As there are no polyatomic interferences for lithium (42), good limits of detection (LoD; 4.014μg/l calculated as 3 x SD of 8 replicates blank) and quantitation (LoQ; 13.381μg/l calculated as 10 x SD of 8 replicates blank) were obtained.
Certified ICP-MS standards and LiCl-based standards were used for droplet lithium quantitation. Certified ICP-MS standards (Agilent) of 500ppb (or µg/l) and LiCl of 240 mg/ml were serially diluted in a matrix-matched artificial CSF solution (148mM NaCl, 3mM KCl, 1.2mM MgCl2, 10mM glucose, 1mg/ml bovine serum albumin) (Figure 2) (43).
From the scans, regions of interest (each spot of plasma, CSF or brain homogenate sample) were selected and average counts per second for each residue were measured and calibrated to standard curve constructed from standards on the same slide. Information on the standards and quantitation used for brain sections and droplets can be found in (43) and (44) respectively. All droplets were analysed in technical triplicates.
Note that there was reasonable concordance in the results from the two methods used for measuring lithium in the samples (Figure 4). However, there was a general tendency for LA-ICP-MS to give higher values, a difference which was more obvious for some of the postnatal brain results (Figure 6). It is likely that the explanation for these differences lies in the different extraction methods used.
Experimental design
Randomisation
In acute experiments, littermates were randomly assigned to the untreated control and experimental groups. In long-term experiments, because all pups within the one litter were exposed to lithium via the mother, age-matched separate litters from untreated dams were used as controls.
Blinding of the study
Analysis of plasma, CSF and brain sections by LA-ICP-MS: samples were identified only by code which was decoded at the end of the experiment. Samples for ICP-MS were also number-coded before being processed and only decoded at the end of the process.
Statistical analysis
Statistical differences between levels of lithium in plasma, CSF and brain as well as CSF/plasma and brain/plasma ratios in acute and long-term treatment group at each age were determined using one-way ANOVA with Tukey’s posthoc test for multiple comparisons (Graphpad Prism 9).