Luteinizing hormone (LH) pulsatility is a hallmark of reproductive function. We describe a protocol for remote activation of specific neuronal populations linked to serial automated blood collection. This technique allows timed hormonal modulation, multiplexing, and minimizing manipulation effects on LH levels in conscious freely moving, and undisturbed animals.
Circulating luteinizing hormone (LH) levels are an essential index of the functioning of the hypothalamic-pituitary control of reproduction. The role of numerous inputs and neuronal populations in the modulation of LH release is still unknown. Measuring changes in LH levels in mice is often a challenge since they are easily disrupted by environmental stress. Current techniques to measure LH release and pulsatility require long-term training for mice to adapt to manipulation stress, certain restraint, the presence of the investigator, and working on individual animals, reducing its usefulness for many research questions.
This paper presents a technique to remotely activate specific neuronal populations using Designer Receptor Exclusively Activated by Designer Drugs (DREADDs) technology coupled with automated sequential blood sampling in conscious, freely moving, and undisturbed mice. We first describe the stereotaxic surgery protocol to deliver adeno-associated virus (AAV) vectors expressing DREADDs to specific neuronal populations. Next, we describe the protocol for carotid artery and jugular vein cannulation and postsurgical connection to the CULEX automated blood sampling system. Finally, we describe the protocol for clozapine-N-oxide intravenous injection for remote neuronal activation and automated blood collection. This technique allows for programmed automated sampling every 5 min or longer for a given period, coupled with intravenous substance injection at a desired time point or duration. Overall, we found this technique to be a powerful approach for research on neuroendocrine control.
The hypothalamo-pituitary-gonadal (HPG) axis is centrally regulated by the pulsatile release of gonadotropin-releasing hormone (GnRH) into the pituitary portal system. In the pituitary gland, GnRH controls the pulsatile release of gonadotropins, luteinizing hormone (LH), and follicle-stimulating hormone (FSH) to the circulatory system. LH pulsatile release serves as a hallmark for central HPG axis functioning1,2,3,4. For example, it displays the effects of genetic alterations or changes in hormonal or environmental factors on the neural part of the axis5,6,7. Until recently, measuring the LH pulsatile pattern was limited to large mammals8 and rats9, given the high frequency of sampling and large blood volumes necessary to identify the pulses.
Detecting LH pulses in mice is desirable since this species has broad genetic models available and can be readily manipulated using genomic engineering technologies to further study specific genes and cell populations. In the last decade, a large advance in the analysis of LH concentrations in mice using a sandwich LH enzyme-linked immunosorbent assay (ELISA) has allowed for detecting LH in a minute amount of blood10. The development of the frequent tail-tip blood sampling technique has made possible the necessary frequent sampling for detecting the frequency and amplitude of LH pulses in mice10,11. However, tail-tip blood sampling is limited to its use in conscious awake animals; it demands a long training period for mice to adapt to the handling and the presence of a designated investigator during sampling. Its success is highly susceptible to environmental stressors and may not be suitable for use in mouse strains showing high levels of anxiety. Intra-atrial cannulation has also been used for frequent blood sampling in freely-moving conscious mice12. However, that setup still requires repeated manual blood sampling and restricts animal moving space, while atrial cannulation can lead to dynamic changes in cardiac function. It is therefore desirable to establish a method for blood collection under stress-free conditions in conscious, freely moving, and undisturbed mice without the need for prior training or human handling or presence.
Automated blood or dialysate sampling has been used before for measuring different hormone levels (e.g., melatonin13,14) and their pulsatile secretion (e.g., growth hormone)15 in unrestrained rodents. We herein present a protocol for automated long-term frequent blood sampling in conscious and unrestrained animals, coupled with a timely remote activation of specific neuronal populations using chemogenetic technologies: the designer receptors exclusively activated by designer drugs (DREADDs). We will describe the stereotaxic delivery of an adeno-associated virus (AAV) vector and the remote activation by an automated intravenous (IV) delivery of clozapine-N-oxide (CNO)16,17. This protocol allows for the sequential detection of basal levels and induced changes in LH pulsatility in multiple animals at the same time. Both the blood sampling and the IV delivery of the compound are conducted in a time-controlled manner via a computer program eliminating the physical presence of the investigator or the requirement for prior mouse training. This method overcomes the main limitations of manual blood sampling. It allows for blood sampling in a stress-free condition, and simultaneous IV compound delivery coupled with remote neuronal activity control. We show representative results of the use of automated blood sampling alone or combined with remote neuronal activation and discuss its advantages, limitations, and additional uses.
All animal procedures are performed in agreement with the National Research Council Guide for the Care and Use of Laboratory Animals18 as well as federal, state, and local laws. Adult female mice (3-6 months old) were used for this protocol demonstration, including four C57BL/6J females and four Kiss1-Cre;ChR2-eYFP (Kiss1-eYFP) females. Mice were held under a 12:12 light:dark cycle, temperature-controlled at 22 °C, and fed ad libitum on a low-phytoestrogen diet. Procedures and protocols were approved by the University of Michigan Institutional Animal Care and Use Committee (IACUC, Animal Protocols: PRO00010420 and PRO00010138).
1. Stereotaxic delivery of AAVs to a specific cell population
2. Jugular vein and carotid artery canulation
3. Automated blood collection and intravenous injection
4. Animal perfusion and brain collection (OPTIONAL)
NOTE: This procedure is to be followed only if the brain is required to analyze the brain site of neuronal activation or downstream responses.
5. Sample processing and analysis
Kisspeptin-expressing neurons (Kiss1 gene) located in the arcuate nucleus of the hypothalamus are a potent stimulator of GnRH, and thus, of LH release from the pituitary gland24,25. In this protocol demonstration, we have used kisspeptin-induced LH secretion to illustrate the functioning of the automated blood sampling technique. Figure 2 shows representative LH patterns in adult Kiss1-eYFP females that previously received a unilateral stereotaxic injection of AAV-hM3Dq-mCherry in the arcuate nucleus (AP: -4.95, ML: -0.35, DV: -5.7). The ChR2-eYFP was used as a fluorescent reporter for Kiss1 cells. One month after the stereotaxic surgery, the mice underwent carotid artery and jugular vein cannulation and were connected to the automatic blood sampler 4 days post-surgery. Blood collections for determining basal LH levels were started on the next day at 10 min sampling frequency (7 min interval between samples and 3 min/sampling) followed by an automated IV CNO injection and continued blood sampling every 10 min for 30 min. Diestrous LH levels are generally low (Figure 2A), but variations are usually observed because of its pulsatile release (Figure 2B). Following the CNO injection (and activation of kisspeptin neurons), the rise in LH was sharp (within 10 min). The level of increase and the duration of the peak depends on numerous factors, including the site of the injection, the number of neurons activated, or the population targeted.
Figure 3 shows the brain site of injection in the arcuate nucleus of the female mouse represented in Figure 2A. Kiss1-eYFP neurons are labeled in green, while the mCherry immunoreactivity shows the site of the AAV injection and activation following CNO. Activated neurons were detected using cFOS immunoreactivity, labeled with DAB. Most mCherry neurons colocalized with Kiss1-eYFP, and many showed cFOS immunoreactivity, demonstrating that the viral and neuronal activation was specific to the targeted population.
A representative LH pulsatile pattern of release in diestrous wildtype (C57BL/6J) mice followed by the response to an IP injection of kisspeptin-10 is shown in Figure 4. The mouse underwent carotid artery cannulation and was connected to the automatic blood sampling system 4 days after the surgery. The next morning, the estrous cycles were checked and the blood collection and kisspeptin-10 injection were performed on the diestrous day26. Blood samples were collected every 7 min for 2 h (4 min interval, plus 3 min/sampling for 120 min before the injection) to determine the baseline levels and LH pulsatility, followed by an IP injection of kisspeptin-10 (65 µg/kg) and continued blood collection every 7 min for an additional 30 min. Clear LH pulses typical for a female in diestrous were observed, showing low basal LH levels, pulse frequency of ~2 pulses/h, and pulse amplitude ~1 ng/mL27. An immediate and robust increase in LH was detected in response to kisspeptin administration28. The LH secretion patterns and changes after stimulation are in agreement with other studies using manual blood collection10,27,29,30. These results demonstrate that the automated blood sampling method captures typical and stimulated LH secretion under a stress-free condition.
Figure 1: Details of the connector system and mouse connections to the infusion and sampling system tubing. (A) A silicone-coated tubing connector (MASA) constructed using two 25G needle tubing and two PE-20 tubing, sleeved in a silastic tubing attached with a small metal ring. (B) A mouse connected to the infusion and sampling tubing in the sampling cage, resting in its nest during the blood sampling. Please click here to view a larger version of this figure.
Figure 2: Representative results for LH pulses in Kiss1-Cre female mice injected with AAV-hM3Dq in the arcuate nucleus and remotely activated with CNO. Basal LH levels were measured every 10 min for half an hour. At time 0 following the blood collection, the female received an intravenous injection of clozapine-N-oxide (0.5 mg/kg) and blood continued to be collected from the carotid artery every 10 min for an additional half an hour. (A) Shows a female with low basal LH levels. (B) Shows a female displaying an LH pulse preinjection. Abbreviations: Kiss1 = Kisspeptin; AAV = adeno-associated virus; CNO = clozapine-N-oxide; LH = luteinizing hormone. Please click here to view a larger version of this figure.
Figure 3: Brain activation in the arcuate nucleus of the Kiss1-Cre;Chr2-eYFP (Kiss1-eYFP) female represented in Figure 2A. ChR2-eYFP was only used as a reporter gene to label kiss1 neurons. (A) Low-magnification fluorescence image showing the site of the AAV injection in the arcuate nucleus. Green: eYFP immunoreactivity, Magenta: mCherry immunoreactivity. (B) Low-magnification brightfield image of the area corresponding to Figure 3A, showing cFOS immunoreactivity (black) in the site of the AAV injection in the Arcuate nucleus. (C) High-magnification fluorescence and brightfield image combined showing a closer look of the neurons in Figure 3A,B. Kiss1-eYFP neurons coexpressing AAV-mCherry that have been activated are those with white cytoplasm and black nucleus (arrows). Scale bars = 50 µm (A,B), 20 µm (C). Abbreviations: Kiss1 = Kisspeptin; AAV = adeno-associated virus; eYFP = enhanced yellow fluorescent protein. Please click here to view a larger version of this figure.
Figure 4: Basal LH pulsatility in a diestrous wildtype female measured every 7 min for 2 h. The female then received an intraperitoneal injection of kisspeptin-10 (65 µg/kg) at time 0, and blood samples were continuously collected every 7 min for half an hour. Abbreviation: LH = luteinizing hormone. Please click here to view a larger version of this figure.
Using this protocol, we were able to show basal LH pulsatility and LH secretion after stimulation of a neuronal population. The great advantages of the system are the stress-free environment in which sampling takes place, with no human presence or handling during the blood sampling. Additionally, no prior laborious animal training and adaptation to human presence or handling during the experiment were required. Previous experiments using manual blood sampling required a large amount of time and effort to minimize stressors7,31,32. However, cutting the tail alone is a stressor33. Implementing a non-stressful environment and training paradigm in shared animal facilities, where interruptions are unpredictable, may also be a constraint. In some laboratories, animals often need to be transported to alternative procedure rooms for blood sampling. These limitations may make the manual method inappropriate for the detection of subtle changes in LH levels, and therefore, a hands-off approach can be helpful in these situations. Automated sampling is set in a quiet room where the mice are placed several days in advance to acclimate to the new environment. Our previous experience with this protocol provided precise detection of corticosterone and pulsatile growth hormone secretion patterns in mice, showing no elevated corticosterone levels during the automated sampling15. In the current experiments, all animals were well adapted to the sampling system showing nest-building in the sampling chamber after ~24 h and bright hair color, indicating a lack of stress and overall good health state (Figure 1).
The main difficulty leading to negative results is probably the inappropriate targeting of the AAV to the required neuronal population. Precision in the stereotaxic injections is essential and training should be done in advance to verify the coordinates and injection volumes. Training can be done by injecting a small amount of 0.5-1% Evans Blue to the desired location in non-recovery surgery and then taking a slice of the freshly dissected brain using a mouse brain matrix (e.g., Ted Pella) to check the site and size of the injection using a stereoscope.
It is also important to take into account that blood and plasma collected from the automated blood sampling system will be diluted in heparinized saline (e.g., 20 µL of blood in 50 µL of saline in our results)20, and the dilution ratio may need to be adjusted for the sensitivity of the selected analytical method. We tested LH levels in whole blood diluted in BSA-PBS (as recommended for Ultra-Sensitive LH ELISA)10 or saline and found no differences in LH values. Tween cannot be used in the diluent since this will circulate into the blood system to extract the samples by replacing the sample fluids20. In our experience, dilutions less than 1:10 gave good LH results but slightly underestimated LH levels as compared to 1:3.5. This indicates that the dilution may be further adjusted to reduce the amount of blood collected, if necessary.
An alternative to automated compound delivery is to do manual injections via the venous catheter. In this case, the investigator is briefly present in the room to deliver the injection. However, there is no direct contact with the animals or their housing and surroundings and, unlike intraperitoneal or subcutaneous injections, the entire procedure is often unnoticed by the animal. The advantages of a manual injection are that the compound dilution does not need to be set up in advance, which may be critical for compounds that are too expensive to use in larger volumes or are sensitive to degradation over time; since the operating volume is smaller than in automated delivery, where the infusion line and catheter need to be prefilled with more compound solution.
Automatic blood sampling provides a unique opportunity to study LH variations during sleep for instance. We have regularly observed animals sleeping in their nests during sampling time. It is possible to link this sampling with EEG recordings to generate a more detailed analysis of the relationship between neural activity and LH pattern34. As shown here, the possibilities for using automated blood sampling are many: from basal LH sampling to testing the LH response to endogenous or exogenous compounds, or to activation or suppression of neuronal populations. The neuronal manipulations can be implemented acutely with chemogenetics or optogenetics, or permanently using transgenic mouse models and apoptotic or neuronal silencing tools. Automated blood sampling also allows for the measurement of other hormones with highly pulsatile secretory patterns (e.g., growth hormone15). In female mice, if a specific phase of the estrous cycle is required, vaginal smears can be carefully collected hours before starting the protocol26 without disrupting the infusion and sampling lines. Animals can be connected to the sampling system for 7-10 days with the risk of clotting the arterial line increasing with time.
This technique, however, is limited to use in single-housed animals, and therefore, it may not be suitable for studying social interactions. It is also invasive and requires technically challenging surgery, so it may not be possible to implement with juvenile animals or certain disease models. Finally, because the cost to purchase the system may be too high for a single research laboratory, it would be advisable to set it up in a core laboratory providing the mentioned experimental protocol as services.
In conclusion, this protocol shows how to perform stereotaxic delivery of AAV combined with automated blood sampling. The precise spatial and temporal control achieved with this technique, together with its flexibility of application to different models, measurement protocols, and hormones, makes it a powerful method for the study of hormonal regulation in rodents. Most importantly, the method provides a stress-free environment by eliminating human presence and handling during injection and/or sampling, and prior animal training. These advantages, together with the possibility of multiplexing, make this method a unique tool for studying the neural control of hormonal changes in conscious, freely moving, and undisturbed mice.
The authors have nothing to disclose.
We thank Dr. Daniel Haisenleder for his help in testing different blood dilution methods. Serum hormone assays were performed at the University of Virginia Center for Research in Reproduction Ligand Assay and Analysis Core supported by the Eunice Kennedy Shriver NICHD Grant R24 HD102061. The Michigan Mouse Metabolic Phenotyping Center-Live is supported by NIH Center Grant U2C DK135066. JF and NQ are supported by DK020572 (MDRC) and DK089503 (MNORC) grants. CFE and CSM are supported by NICHD grant R21 HD109485 and R01 HD096324.
AAV8-hSyn-hM3D(Gq)-mCherry | Addgene | 44361 | Not necessarily this virus but this was the one used for representative results |
Alcohol | Disinfection | ||
Anesthesia Induction box | Vetequip | ||
Anesthesia induction machine | Kent Scientific Equipment | SomnoSuite | |
Anesthesia masks for mice | Kent Scientific Equipment | SOMNO-0801 | |
Autoclip applier 9 mm | Clay Adams | 427630 | |
Autoclip remover 9 mm | Clay Adams | 427637 | |
Autoclips 9 mm | Clay Adams | 427631 | |
BASi Culex Controller | Culex | SN: 2151, 2152, 2156, 2158 | 4 stations |
BASi Honey Comb Fraction Collector | Honey Comb | SN: 2105, 2106, 2107, 2108 | 4 stations |
BASi Ratrun Rotation Control | RATURN 2 | SN: 5680, 5681, 5682, 5683 | 4 stations |
C57BL/6J mice | JAX # 000664 | ||
Carprofen | Zoetis | Rimadyl | Analgesic |
Clippers | Braun | ||
Clozapine-N-oxide | ENZO | BLM-NS105-0005 | |
Cotton tipped applicators | |||
CULEX Automated In Vivo Sampling System | BASi | DS000627 | with CX-4000S Replacement Tubing Sets |
Curved forceps serrated | FST | 11151-10 | |
Drill | Dremel | 61100 | |
Empis control Module | EMPIS CM | SN: 174 | |
Empis Programmable Infusion System | EMPIS | SN: 2125 , 2126, 2127, 2128 | With CX-7010S 4 BAS-2 Infusion Sets; 4 stations |
Envigo 2016 diet | low-phytoestrogen diet | ||
Eye ointment | Dechra | Puralube Vet Ointment | Petrolatum Ophtalmic oinment |
Glass pipettes | World Precision Instruments | MIB100-6 | |
Hemostats | Roboz Surgical | RS-7101 | |
Iodine | Betadine Surgical scrub | ||
Isoflurane | VetOne | Fluriso | Anesthetic |
Isoflurane Vaporizer or SomnoSuite Low-Flow Anesthesia System | Surgivet or Kent Scientific Corp | SS-01 | Anesthesia Machine |
Kiss1-Cre;ChR2-eYFP (Kiss1-eYFP) mice | JAX # 023436 and #024109 | ||
Kisspeptin-10 | Phoenix Pharmaceuticals | 048-56 | |
Micro-renathane tubing | Braintree Scientific | MRE025 | Surgical catheterization |
Micro-Scissors | Roboz Surgical | RS-5606 | |
Needle Holder | Roboz Surgical | RS-7842 | |
Picoliter injector | Warner Instruments | PLI-100A | |
Pipette puller | Sutter Instruments | P30 | |
Rodent Warmer X2 | Stoelting | 53850 | |
Scalpel | FST | 10003-12 | |
Scissors | Roboz Surgical | RS-6808 | |
Silicon tubing | Liveo Laboratory Tubing | NO.508-001 | 0.012 in I.D x 0.025 in O.D. |
Stereotaxic table | RWD | E06208 | |
Sterile 0.9% saline | Baxter | 2F7124 | |
Sterile towel drapes | Dynarex | 4410 | |
Surgical blades | SKLAR | 06-3011 | |
Surgical stereoscope | Zeiss | f-160 | |
Tweezers | Roboz Surgical | RS-4960 | |
Tweezers | Roboz Surgical | RS-4972 | |
Tweezers | Roboz Surgical | RS-5058 | |
Antibodies | |||
Anti-cFos | Millipore | ABE457 | Antigen target: N-terminus cFos; Host organism: Rabbit; Dilution used: 1:5,000; RRID: AB_2631318 |
Anti-GFP | Aves Labs | GFP-1010 | Antigen target: recombinant GFP null; Host organism: Chicken; Dilution used: 1:10,000; RRID: AB_2307313 |
Biotin-SP-conjugated AffiniPure Donkey Anti-Rabbit IgG | Jackson ImmunoResearch Labs | 711-065-152 | Antigen target: Rabbit IgG (H+L); Host organism: Donkey; Dilution used: 1:1,000; RRID: AB_2340593 |
Donkey anti-Rat IgG, AlexaFluor 594 | Thermo Fisher Scientific | A-21209 | Antigen target: Rat IgG (H+L); Host organism: Donkey; Dilution used: 1:500; RRID: AB_2535795 |
Goat anti-Chicken IgY, Alexa Fluor 488 | Thermo Fisher Scientific | A-11039 | Antigen target: Chicken, IgY (H+L); Host organism: Goat; Dilution used: 1:500; RRID: AB_2534096 |
mCherry monoclonal (16D7) | Thermo Fisher Scientific | M11217 | Antigen target: mCherry tag; Host organism: Rat; Dilution used: 1:5,000; RRID: AB_2536611 |