Lithium sulfate

Using the protocol described before [16 (link)], 24 h before the experiments, the rats were exercised by 20 consecutive front and hind paw grips on a vertical grid, gently held at the base of their tail as previously described [54 (link)].
The same procedure was repeated on 3 consecutive days. Immediately after administration of lithium sulfate (500 mg/kg ip), rats received 10 µg/kg BPC 157, 10 ng/kg BPC 157, or 5 mL/kg saline intraperitoneally, and after 20 min, the rats were placed on an upside-down grid. If the rats fell, they were continuously placed again within 1 min until the end of the 8 min period. At 3 h after the end of each session, the rats were euthanized.
Recording of the brain swelling in rats was performed 15 min after a complete calvariectomy. Briefly, six burr holes were drilled in three horizontal lines, all of them medial to the superior temporal lines and temporalis muscle attachments. The two rostral burr holes were placed just basal from the posterior interocular line, the two basal burr holes were placed just rostral to the lambdoid suture (and transverse sinuses) on both sides, and the middle two burr holes were placed in the line between the basal and rostral burr holes. The procedure was undertaken 3 h after each of the sessions. Alternatively, lithium administration was recorded as time 0, and brain presentation was recorded in healthy rats 15 min before lithium administration (−15 min → 0). Then, lithium 500 mg/kg in saline was given intraperitoneally (time 0), and the 0 → +3 min period was recorded. Thereafter, the +3 min (5 mL/kg saline intraperitoneally) → +6 min period was recorded. Then, the administration of BPC 157 was recorded during the next 3 min period (+6 min → +9 min period).
A laparotomy was performed for the corresponding presentation of the peripheral veins (superior mesenteric, inferior caval, and azygos veins), and a camera attached to a VMS-004 Discovery Deluxe USB microscope (Veho, Dayton, OH, USA) was used for recording. The procedure was undertaken 3 h after each of the sessions.
Muscular weakness. The amount of time the rats could hold on to the grid reflected the grade of fatigue and muscle weakness. As described previously, the scoring assessment (0–5, healthy rats presented with scores of 4 and 5) was carried out in 1 min intervals until the end of the session as follows: 0/5: immediately falling, no contraction, hunched posture with flaccid paralysis; 1/5: falling (\5 s), muscle flicker, but no movement, hunched posture upon falling; 2/5: falling (\10 s), movement possible, but not against gravity, hunched posture upon falling; 3/5: falling (\20 s), movement possible against gravity, but not against resistance by the examiner, hunched posture upon falling; 4/5: no obvious fatigue, movement possible against some resistance by the examiner, normal posture and activity upon falling (150 s); 5/5: no fatigue, movement possible against significant resistance by the examiner, normal posture and activity upon falling (150 s).
ECG recording. ECGs were recorded continuously in deeply anesthetized rats for all three main leads by positioning stainless steel electrodes on all four limbs using an ECG monitor with a 2090 programmer (Medtronic, Minneapolis, MN, USA) connected to a Waverunner LT342 digital oscilloscope (LeCroy, Chestnut Ridge, NY, USA) prior to sacrifice. This arrangement enabled precise recordings, measurements, and analyses of ECG parameters [2 (link),3 (link),4 (link),5 (link),6 (link),7 (link),8 (link),9 (link)]. The procedure was undertaken at 3 h after each of the sessions.
Thrombus assessment. On being euthanized, the superior mesenteric vein and superior mesenteric artery were removed from the rats, and the clots were weighed [2 (link),3 (link),4 (link),5 (link),6 (link),7 (link),8 (link),9 (link)].
Superior sagittal sinus, portal and inferior caval veins, and abdominal aorta pressure recordings. As described previously [2 (link),3 (link),4 (link),5 (link),6 (link),7 (link),8 (link),9 (link)], recordings were made in deeply anesthetized rats with a cannula (BD Neoflon™ Cannula, BD Switzerland, Eysins, Switzerland) connected to a pressure transducer (78534C MONITOR/TERMINAL; Hewlett Packard, Palo Alto, CA, USA) that was inserted into the superior sagittal sinus, portal and inferior caval veins, and abdominal aorta at the level of the bifurcation at 3 h after each session, after 5 min of recording. For the superior sagittal sinus pressure recordings, we made a single burr hole in the rostral part of the sagittal suture just above the superior sagittal sinus and cannulated the anterior portion of the superior sagittal sinus with Braun intravenous cannulas. Then, we laparatomized the rats to cannulate the portal vein, inferior caval vein, and abdominal aorta for their respective pressure recordings.
Notably, normal rats exhibited a superior sagittal sinus pressure of −24 to −27 mmHg and portal pressure of 3–5 mmHg, which was similar to that of the inferior caval vein, although with at least 1 mmHg higher values in the portal vein. By contrast, the abdominal aorta blood pressure values at the level of the bifurcation were 100–120 mm Hg [2 (link),3 (link),4 (link),5 (link),6 (link),7 (link),8 (link),9 (link)].
Brain volume and vessel presentation. Brain volume and vessel presentation were proportional with the change in the surface area of the brain or vessel. We used the protocol previously described [4 (link),6 (link),7 (link),8 (link),9 (link)]. At 3 h after each of the sessions, the presentations of the brain and peripheral veins (superior mesenteric, inferior caval, and azygos veins) were recorded in deeply anaesthetized rats, with a camera attached to a VMS-004 Discovery Deluxe USB microscope (Veho, Dayton, OH, USA), before the procedure in control rats and just before sacrifice in rats administered lithium. The borders of the brain or veins in the photographs were marked using ImageJ computer software. Then, the surface area (in pixels) of the brain or veins was measured using a measuring function. This was performed with brain photographs before the application and at intervals after the application for both control and treated animals. In the rats with occluded mesenteric veins, the surface area of the brain or vein before application was marked as 100%, and the ratio of each subsequent brain area to the first area was calculated ( $$\frac{{\mathrm{A}}_2}{{\mathrm{A}}_1}$$ ). Using the square-cube law shown in Equations (1) and (2), an equation for the change in brain volume proportional to the change in the surface area of the brain (6) was derived. In expressions (1)–(5), l is defined as any arbitrary one-dimensional length of the brain (for example, the rostro-caudal length of the brain) and is used only to define the one-dimensional proportion (l₂/l₁) between two observed brains and as an inter-factor (and, therefore, was not measured [6 (link)]) to derive the final expression (6). The procedure was as follows: $${\mathrm{A}}_2={\mathrm{A}}_1\times {\left(\frac{{\mathrm{l}}_2}{{\mathrm{l}}_1}\right)}^2$$
(Square-cube law), $${\mathrm{V}}_2={\mathrm{V}}_1\times {\left(\frac{{\mathrm{l}}_2}{{\mathrm{l}}_1}\right)}^3$$ ; (2) (Square-cube law), $$\frac{{\mathrm{A}}_2}{{\mathrm{A}}_1}={\left(\frac{{\mathrm{l}}_2}{{\mathrm{l}}_1}\right)}^2$$ ; (3) (from (1), after dividing both sides by A₁), $$\frac{{\mathrm{l}}_2}{{\mathrm{l}}_1}=\sqrt{\frac{{\mathrm{A}}_2}{{\mathrm{A}}_1}}$$ ; (4) (from (3), after taking the square root of both sides), $$\frac{{\mathrm{V}}_2}{{\mathrm{V}}_1}={\left(\frac{{\mathrm{l}}_2}{{\mathrm{l}}_1}\right)}^3$$ ; (5) (from (2), after dividing both sides by V₁), $$\frac{{\mathrm{V}}_2}{{\mathrm{V}}_1}={\left(\sqrt{\frac{{\mathrm{A}}_2}{{\mathrm{A}}_1}}\right)}^3$$ ; and (6) (after incorporating expression (4) into Equation (5)).
Stomach lesions. The presentations of the gross lesions in the gastrointestinal tract were recorded in deeply anaesthetized rats, with a camera attached to a VMS-004 Discovery Deluxe USB microscope (Veho, USA). At 3 h after each of the sessions, we assessed the hemorrhagic congestive lesions in the stomach (sum of the longest diameters (in mm)).
Microscopy. Tissue preparation. The brain, liver, kidney, lungs, heart, stomach, intestines, and quadriceps muscle tissues were fixed in 10% neutral buffered formalin (pH 7.4) at room temperature for 24 h. Representative tissue specimens were dehydrated and embedded in paraffin, sectioned at 4 μm, and stained with hematoxylin-eosin according to the following automated Sakura Tissue-Tek DRS 2000 Slide Stainer protocol (https://www.sakura.eu/Solutions/Staining-Coverslipping/H-E-Kit accessed on 19 October 2021): rehydration in distilled water, staining with hematoxylin, washing in running tap water, differentiation with 70% alcohol, staining with eosin, dehydration, clearing, and mounting. Tissue injury was evaluated microscopically by two blinded examiners (board-certified pathologists, A.S. and E.L.) using an Olympus BX51 microscope and an Olympus 71 digital camera for saving images as uncompressed 24-bit RGB TIFF files.
Brain histology. Brain injury in different regions [4 (link),6 (link),7 (link),8 (link),9 (link),55 ] was evaluated using a semiquantitative neuropathological scoring system as described [4 (link),6 (link),7 (link),8 (link),9 (link),56 (link)] (Table 1), providing a common score 0–8 (grade 0 indicates no histopathologic damage).
Lung histology. A scoring system to grade the degree of lung injury was used in lung tissue analysis. Features were focal thickening of the alveolar membranes, congestion, pulmonary edema, intra-alveolar hemorrhage, interstitial neutrophil infiltration, and intra-alveolar neutrophil infiltration. Each feature was assigned a score from 0 to 3 based on its absence (0) or presence to a mild (1), moderate (2), or severe (3) degree, and a final histology score was determined [4 (link),6 (link),7 (link),8 (link),9 (link),57 (link)].
Renal, liver, and heart histology. The criteria for renal injury were based on degeneration of Bowman’s space and glomeruli, degeneration of the proximal and distal tubules, vascular congestion, and interstitial edema. The criteria for liver injury were vacuolization of hepatocytes and pyknotic hepatocyte nuclei, activation of Kupffer cells, and enlargement of sinusoids. Each specimen was scored using a scale ranging from 0 to 3 (0: none, 1: mild, 2: moderate, and 3: severe) for each criterion, and a final histology score was determined [4 (link),6 (link),7 (link),8 (link),9 (link),58 ]. Myocardial injury features used in analyzing heart lesions were based on the severity of congestion (each specimen was scored using a scale ranging from 0 to 3 (0: none, 1: mild, 2: moderate, and 3: severe), and a final histology score was determined) and the presence or absence of myocardial infarction.
Intestinal histology. A histologic scoring scale adapted from Chui et al. [4 (link),6 (link),7 (link),8 (link),9 (link),59 (link)] was used for tissue scoring on a scale of 0 to 5 (normal to severe) in three categories (mucosal injury, inflammation, and hyperemia/hemorrhage) for a total score of 0 to 15, as described by Lane et al. [4 (link),6 (link),7 (link),8 (link),9 (link),60 (link)]. The morphologic features of mucosal injury were based on different grades of epithelia lifting, villi denudation, and necrosis; grades of inflammation were graded from focal to diffuse according to lamina propria infiltration or subendothelial infiltration; and hyperemia/hemorrhage was graded from focal to diffuse according to lamina propria or subendothelial localization.
Muscle histology. Transverse sections of the quadriceps muscle were used for histological evaluation and examined in a blinded fashion. A special software program, ISSA Network Station Version 4.0. (VAMSTEC, Zagreb, Croatia), was used for morphometric analysis. Five high-power fields from the quadriceps muscle, which were examined as semi-serial muscle sections, were randomly selected for analysis. In the selected areas, the smallest diameters of the smallest muscle fibers were measured as previously described, and the healthy values of the quadriceps muscle (31 ± 3 mm) were considered normal [53 (link),61 (link),62 (link),63 (link)].
Oxidative stress. At the end of the experiment, at 3 h after each session, oxidative stress in the collected tissue samples (brain, heart, lung, liver, kidney, and quadriceps muscle) was assessed by quantifying the thiobarbituric acid-reactive species (TBARS) as malondialdehyde (MDA) [34 (link),35 (link),36 (link)]. The tissue samples were homogenized in PBS (pH 7.4) containing 0.1 mM butylated hydroxytoluene (BHT) (TissueRuptor, Qiagen, Valencia, CA, USA) and sonicated for 30 s in an ice bath (Ultrasonic Bath, Branson, MI, USA). Trichloroacetic acid (TCA, 10%) was added to the homogenate, the mixture was centrifuged at 3000 rpm for 5 min, and the supernatant was collected. Then, 1% TBA was added, and the samples were boiled (95 °C, 60 min). The tubes were then kept on ice for 10 min. Following centrifugation (14,000 rpm, 10 min), the absorbance of the mixture was determined at the wavelength of 532 nm.
The concentration of MDA was read from a standard calibration curve plotted using 1,1,3,3-tetraethoxypropane (TEP). The extent of lipid peroxidation was expressed as MDA using a molar extinction coefficient for MDA of 1.56 × 10⁵ mol/L/cm. The protein concentration was determined using a commercial kit. The results are expressed in nmol/g of protein.
Lithium analysis. At 20 min or 3 h after administration of lithium sulfate (500 mg/kg ip) (lithium-time), followed by BPC 157 10 ng/kg ip, or saline 5 mL/kg ip, the rats were euthanized and blood and tissue samples taken for lithium analysis, i.e., serum, brain, muscle, heart, stomach, spleen, kidney, liver, and intestine. Blood samples were collected in vacutainer tubes without anticoagulant (BD Vacutainer Trace Element Serum, Ref 368380) (Becton-Dickinson, Franklin Lakes, NJ, USA), centrifuged at 3000 rpm for 15 min, and serum transferred into 2 mL CryoPure Tubes (Sarstedt, Nümbrecht, Germany) was kept at −20 °C until analysis. Lithium in serum and tissue samples was quantified by inductively coupled plasma mass spectrometry (ICP-MS) using Agilent 7500 cx (Agilent Technologies, Tokyo, Japan). Tissue samples were prepared for the analysis by microwave-assisted digestion in 75% (v/v) HNO₃ in an UltraCLAVE IV (Milestone, Italy) microwave following the procedure detailed elsewhere (Vihnanek Lazarus 2013). After digestion, samples were adjusted to 6 g with ultrapure water (GenPure, TKA System GmbH, Niederelbert, Germany), and additionally diluted to 1:10 with 1% (v/v) HNO₃ and 3 µg/L internal standards before analysis. Serum samples were diluted 1:20 with a solution containing 0.7 mM ammonia, 0.01 mM EDTA, 0.07% Triton X-100, and 3 µg/L of internal standards in ultrapure water. The blanks, matrix-matched calibration standards, and reference materials were prepared in the same manner as the samples. Single-element standard solutions (1000 ± 7 mg/L) used for calibration (Li) and as internal standards (Ge, Rh, Tb and Ir, Lu) were obtained from SCP SCIENCE (SCP Science, QC, Canada). To confirm the accuracy of the measurements, the following reference materials in serum/plasma were used: ClinChek^® Serum Controls (Levels I and II), ClinChek^® Plasma Controls (Levels I and II), (Recipe, Munich, Germany), and SeronormTM Trace Elements Serum (Levels I and II), (Sero AS, Billingstad, Norway). Overall recoveries were in the range of assigned analytical values. Previously reported concentrations of lithium in the rat plasma from 3 month old female and male Wistar rats were 6.5 ± 0.73 µg/L and 5.8 ± 0.28 µg/L, respectively [64 (link)].

Radiolabeled and chromogenic HIV-1 and SIV in situ hybridization (R-ISH and C-ISH, respectively) were performed as previously described with some modifications [15 (link), 17 (link)]. In brief for R-ISH, after acetylation with acetic anhydride and dehydration, slides were hybridized at 45°C overnight with a ³⁵S-labeled riboprobe containing 0.5 mM aurintricarboxylic acid in the hybridization mix. After extensive washes and ribonuclease treatment, tissue sections were dehydrated, coated in Ilford K5 or Kodak NTB emulsion diluted with glycerol and ammonium acetate, exposed at 4°C for 11 days, and developed and fixed per manufacturer's instructions. Slides were counterstained with hematoxylin, dehydrated, and mounted with Permount.
Chromogenic HIV-1 and SIV in situ hybridization (C-ISH) was performed as previously described [17 (link)]. In addition, we utilized RNAscope for SIV and HIV-1 detection [50 (link)]. Each series of specific SIV and HIV-1 target probes was designed to hybridize to viral RNA in gag, pol, vif, vpr, tat, rev, env, nef, and vpx genes (for SIV) (Supplemental Table 2). In addition to probes for SIV and HIV-1 clade B described in this report, we have generated and validated RNAscope probes for HIV-1 clades A, CRF_AE, C, and D with similar results to SIV (Supplemental Table 2; data not shown). The target probe design strategy was described previously [19 (link)]. Briefly, each target probe contained an approximately 25-base region complementary to the corresponding SIV (or HIV-1) plus-RNA strand (transcribed transcripts or whole transcribed genome) of each gene, a spacer sequence, and a 14-base tail sequence (conceptualized as a Z). A pair of target double Z probes (ZZ), each possessing a different type of tail sequence, hybridized contiguously to a target region (approximately 50 bases). The two-tail sequences together formed a 28-base hybridization site that binds to a signal preamplifier, which initiates a signal amplification cascade via sequential hybridization, similar to the branched DNA (bDNA) method described previously [20 (link)], followed by chromogenic enzymatic detection (horseradish peroxidase using 3, 3-diaminobenzidine [DAB] or tyramide-cyanine 3.5 [PerkinElmer] or alkaline phosphatase using Fast Red substrate). This approach targeted about 4.5 kb of the viral genome. The double-Z probe design strategy ensures superior background control because it is highly unlikely that a nonspecific hybridization event will juxtapose a pair of target probes along an off-target molecule to form the 28-base hybridization site required for binding of the preamplifier and also because a single 14-base tail sequence will not bind the preamplifier with sufficient strength to result in successful signal amplification. Sections of tissues or cell pellets (4-6 μm) were mounted on Superfrost Plus microscope slides (Fisher Scientific), heated at 60°C for 1 hour, dewaxed in xylenes for 10 minutes, and then placed in ethanol 100% for 5 minutes before air drying. RNAscope was performed as previously described [19 (link)]. First, slides were incubated with RNAscope Pretreat 1 reagent (endogenous peroxidase block; ACD) for 10 minutes at room temperature. Heat-induced epitope retrieval was performed by boiling sections in RNAscope Pretreat 2 buffer (a citrate buffer [10 nmol/L, pH 6]; ACD) for 30 minutes, immediately washed in double distilled water, and then dehydrated in 100% ethanol for 5 minutes before air drying. Hydrophobic barrier pen was applied to encircle the section, then the slides were incubated with diluted (1:5) RNAscope pretreat 3 reagent (protease digestion solution; 2.5 ug/mL) for 20 to 25 minutes at 40°C using a HybEZ hybridization oven (ACD). Sections were rinsed 3 times in double distilled water and then incubated with pre-warmed target probes (20 nmol/L of each oligo probe) in hybridization buffer A (6X SSC [1XSSC is 0.15 mol/L NaCl, 0.015 mol/L Na-citrate], 25% formamide, 0.2% lithium dodecyl sulfate, blocking reagents) and incubated for 2 hours at 40°C. Slides were washed in wash buffer (0.1X or 0.05X SSC, 0.03% lithium dodecyl sulfate) and incubated with amplification reagents as described in the RNAscope 2.0 HD detection protocol. Amplifier 1 (2 nmol/L) in hybridization buffer B (20% formamide, 5X SSC, 0.3% lithiumdodecyl sulfate, 10% dextran sulfate, blocking reagents) at 40°C for 30 minutes; Amplifier 2 (a proprietary enhancer to boost detection efficiency) at 40°C for 15 minutes; Amplifier 3 (2 nmol/L) in hybridization buffer B at 40°C for 30 minutes; Amplifier 4 (2 nmol/L) in hybridization buffer C (2X SSC, blocking reagents) at 40°C for 15 minutes; Amplifier 5 (a proprietary signal amplifier) at room temperature for 30 minutes; Amplifier 6 (a proprietary secondary signal amplifier) at room temperature for 15 minutes. After each hybridization step, slides were washed with wash buffer three times at room temperature. Before detection, the slides were rinsed one time in 1X TBS Tween-20 (0.05% v/v). Amplification 6 contained alkaline phosphatase (or horseradish peroxidase) labels, and chromogenic detection was performed using FastRed as substrate to generate red signal, DAB to generate a brown signal, or tyramide-cyanine 3.5 (PerkinElmer) for fluorescence detection. Red chromogen development was performed following the RNAscope 2.0 HD detection protocol and reagents, brown chromogen development using ImmPACT™ DAB (Vector Laboratories), and fluorescent detection using tyramide-cyanine 3.5 Plus (PerkinElmer) (chromogen incubation time varied between 2 and 8 minutes). Slides were counterstained with haematoxylin or DAPI (4′,6-diamidino-2-phenylin-dole) and mounted in Permount (Fisher Scientific) or Prolong^® Gold (ThermoFisher Scientific). Slides mounted in Permount were scanned at high magnification (×400) using the ScanScope AT2 System (Aperio Technologies), yielding high-resolution data from the entire tissue section. Fluorescent slides mounted with Prolong^® Gold (Invitrogen) were imaged on an Olympus FV10i confocal microscope using a 60x phase contrast oil-immersion objective (NA 1.35) and applying a sequential mode to separately capture the fluorescence from the different fluorochromes at an image resolution of 1024 x 1024 pixels.

Prior to the crystallization, the purified CMY-185 recombinant enzyme was diluted to the final concentration of 15 mg/mL in 20 mM Tris-HCl pH 7.0. Crystals of CMY-185 were obtained using hanging-drop vapor diffusion method with the crystallization condition in which the CMY-185 solution was mixed with an equal volume of a reservoir solution containing 24% PEG 20,000, 0.1 M Tris-HCl pH 8.5, and 0.2 M lithium sulfate at 20°C. Crystals of the CMY-185-ceftazidime complex were obtained by soaking the CMY-185 crystals in a solution containing 100 mM ceftazidime, 26% PEG 20,000, 0.1 M Tris-HCl pH 7.0, and 0.2 M lithium sulfate at 4°C for 4 h.

FISH was performed with a FISH kit (Ribo Biotechnology) according to the manufacturer’s instructions. In brief, cells were fixed with 4% PFA (Sigma-Aldrich) for 10 min at RT and permeabilized with 0.5% Triton X-100 (Sigma-Aldrich) for 15 min at RT. Prehybridization was performed with lncRNA FISH probe mix at 37 °C for 30 min, and then hybridization was performed by adding lncRNA FISH probe mix and incubating the mixture at 37 °C overnight. After washing with 4×, 2×, and 1× SSC (1×SSC is 0.15 M NaCl, 0.015 M Na-citrate), the cell nuclei were stained with DAPI (Thermo Fisher).
RNAScope was performed with an RNAscope Fluorescent Multiplex Reagent Kit (ACD Bio) based on the manufacturer’s protocols. In short, cells were first placed on slides and fixed in 4% PFA (Sigma-Aldrich) for 30 min, followed by antigen repair with RNAscope® hydrogen peroxide (ACD Bio) for 10 min at RT and digestion with RNAscope® protease III (ACD Bio) for another 10 min at RT in a humidifying box. Next, the cells were incubated at 40 °C with the following solutions: (1) RNAScope probes of target RNAs, namely, lnc-ip65-C3 and HTNV-S-C2 (v/v, 1:1), in hybridization buffer A (6 × SSC, 25% formamide, 0.2% lithium dodecyl sulfate, blocking reagents) for 2 h; (2) preamplifier (AMP1, 2 nM) in hybridization buffer B (20% formamide, 5×SSC, 0.3% lithium dodecyl sulfate, 10% dextran sulfate, blocking reagents) for 30 min; (3) amplifier (AMP2, 2 nM) in hybridization buffer B at 40 °C for 30 min; and (4) label probe (AMP3, 2 nM) in hybridization buffer C (5×SSC, 0.3% lithium dodecyl sulfate, blocking reagents) for 15 min. After each hybridization step, the slides were washed with wash buffer (0.1×SSC, 0.03% lithium dodecyl sulfate) three times at RT. Then, the probe signaling was further recognized and amplified by HRP-C2 (ACD Bio) (for 15 min at 40 °C), followed by chromogenic detection with TSA® Plus Cy3 (Akoya Biosciences) (for 30 min at 40 °C) for detecting HTNV-S. After treatment with HRP-C2-blocker (ACD Bio), the aforementioned steps were repeated with HRP-C3 (ACD Bio) and TSA® Plus Cy5 (Akoya Biosciences) to assess lnc-ip65. Finally, after DAPI staining and Prolong Gold Antifade Mountant (Thermo Fisher) treatment, the samples underwent IFA for p65 detection or were directly observed with a confocal microscope (Nikon).