An Efficient Single-Person Technique for Milk Sampling from Laboratory Mice

Summary

This article describes how to assemble a device that can be used to collect milk from laboratory mice. This device is affordable, portable, and allows a collection of 0.1–0.5 mL of milk by one trained person. A detailed milking protocol is also included. 

Abstract

Rodent models have been widely used in biology since the early 20th century to study basic physiology and biochemical mechanisms. In lactation and neonatal development research, there is a need to sample milk to determine the composition and how milk may be affected by experimental manipulations, thereby impacting newborn development. Collecting biofluids in sufficient amounts from small rodents can be difficult. Several studies have demonstrated different techniques to obtain milk from mice. However, these techniques often require at least two trained personnel, which may be challenging in some cases. 

Here, we demonstrate a modified milk sampling technique based on a method published in 2009. With this affordable and easy-to-assemble method, one trained person can routinely obtain 0.1–0.5 mL of milk from a dam under anesthesia in less than 10 min. Using this method, we collected sufficient milk on post-delivery day 2 (PD2) and post-delivery day 10 (PD10). We measured milk macronutrient components and compared them with the existing literature to validate our collection method. 

We recorded that on PD2, milk protein averaged at 70.1 ± 5.2 g/L, milk fat was 174.4 ± 7.1 g/L, and milk lactose was 12.3 ± 0.6 g/L. On PD10, while milk lactose and protein remained at similar concentrations as on PD2, milk fat was significantly higher (224.4 ± 17.1 g/L). We also observed that the relative abundance of individual milk proteins varied between PD2 and PD10. More specifically, αS1-casein and whey acidic protein were higher, while β-casein was lower at PD10 compared with PD2. Overall, we demonstrate an efficient single-person technique for milk sampling from mice using a device that can easily be assembled with commonly used laboratory equipment.

Introduction

Comparative medicine refers to the idea that humans and other animals share similarities in anatomy and physiology; thus, information learned from one species can be used to study similar pathways in the other^1,2. Since the early 20^th century, rodents, specifically mice and rats, have been used widely in biomedical research due to relatively easy genetic manipulation for the development of disease models^1,2. Additionally, they are relatively small; thus fewer resources are required to maintain colonies compared with other non-rodent mammalian models². However, their small body sizes also come with challenges. Certain procedures cannot be done easily on mice. For instance, the arteriovenous difference technique can be performed on rats^3,4,5, but it can be challenging in mice as their small blood vessels can easily rupture^6,7. The amount of tissue that can be sampled from a mouse is also limited. For example, in lactating mice, visceral adipose tissue decreases significantly in size^8,9 and in some cases, to almost undetectable amounts. Additionally, the sampling of biofluid from mice is also limited in amount; depending on the technique, the volume of blood sampled may vary but is always somewhat limited⁷, restricting the number of analyses that can be performed.

In lactation and neonatal development research, there is a need to sample milk to evaluate composition and milk yield. In studies performed in mice, due to limited milk volume, samples are often pooled from several dams to yield an adequate amount¹⁰. A number of studies demonstrate techniques for the collection of sufficient analytical volumes of milk from a single mouse^11,12,13. These methods, however, usually require two trained personnel^11,12: one to hold the animal while manually expressing milk and one to collect milk with a pipette; or they require equipment and tools that are not readily available in a laboratory. Here, we demonstrate a modification of a technique devised by DePeters and Hovey¹⁴. The device used in this procedure can be easily assembled with equipment available in the laboratory or easily purchased from common vendors. Using this device, one person can routinely obtain 0.1–0.5 mL of mouse milk in less than 10 min to use for most analyses. We collected milk on post-delivery day 2 (PD2, early lactation) and post-delivery day 10 (PD10, peak lactation) and measured milk macronutrients to validate the method.

Protocol

All animal care and experimental procedures followed federal guidelines and received approval from the Rutgers University Institutional Animal Care and Use Committee. Female mice (C57BL/6 background) were mated at 8 weeks of age. They were fed a standard breeder chow ad libitum and kept under a regular 12:12 h light:dark cycle. The day of delivery was counted as post-delivery day 1 (PD1). On PD2 (early lactation) and PD10 (peak lactation), milk was sampled as described below.

1. Assembly of a milking device

1. Attach one end of the 5/32'' inner diameter (ID) polyvinyl chloride (PVC) tubing to the vacuum pump and use tape to secure the attachment (labeled a in Figure 1A).
   NOTE: The vacuum pump can be substituted with any suitable pump or a central vacuum (sometimes called “house vacuum”) available in the laboratory.
2. Attach the other end of the 5/32'' ID PVC tubing to the 1/8'' ID PVC tubing via a barbed straight connector.
3. Attach the other end of the 1/8'' ID PVC tubing to an 18G hypodermic needle via a connector (labeled b in Figure 1A).
4. Cut the end of a P200 pipette tip ~0.2 cm to create a larger opening. Insert the P200 pipette tip through the opening of the septum stopper so that the tip points upward (Figure 1B).
   NOTE: Change pipette tips in between animals.
5. Attach the septum stopper to a sterile collecting tube. Make a puncture through the septum stopper with the 18 G hypodermic needle.
6. Check the suction pressure by turning on the vacuum pump and placing a finger (covered in a laboratory glove) against the pipette tip opening.
   NOTE: You should feel a gentle suction if it is assembled correctly. The collecting tube, 18 G hypodermic needle, and P200 pipette tips should be sterile.

Figure 1: The milking device. (A) Milking device with a vacuum pump attached to an 18 G hypodermic needle via PVC tubing. The needle goes through a septum stopper that is attached to a microcentrifuge collecting tube. An inverted P-200 pipette tip, which is used to draw milk from a teat, is inserted through an opening of the septum stopper. (B) Cross-sectional diagram of the milking device. (C) A dam being milked using the described method. (D) A representative aliquot (0.3 mL) of milk collected from a dam at peak lactation. Please click here to view a larger version of this figure.

2. Separation of the dam from pups

1. On the day of milking, separate the dam from her pups to allow milk accumulation in the glands.
2. Keep the pups warm in a nest.
3. Allow 2 h of separation if milk is collected during early lactation (e.g., PD2 in this study).
4. Allow 30–45 min of separation if milk is collected during peak lactation (e.g., PD10 in this study).

3. Anesthesia of the dam

1. When the dam is ready to be milked, administer an injection of ketamine/xylazine solution (intraperitoneal, i.p.) into one side of the peritoneal cavity between the 4^th and 5^th pair of teats.
   NOTE: The amount of anesthetic solution to be administered is calculated by the weight of the dam (80–100 mg/kg ketamine, 5–12 mg/kg xylazine, according to Rutgers University IACUC protocol). This dose will sedate a mouse for 45–60 min, providing ample time for milk collection.
2. Confirm anesthesia by squeezing a foot firmly, but without injuring it, to test the animal’s perception of sensation and pain. Adequate anesthesia is indicated by no movement of any kind.
3. Apply veterinary ophthalmic ointment to the eyes of the dam to prevent dryness while under anesthesia.

4. Milk collection

1. Place the anesthetized dam on her back on a solid working surface. Gently wipe her mammary area with a sterile alcohol prep wipe to clean the area prior to milking.
2. Administer 1–2 units of oxytocin (i.p.) between the 4^th and 5^th pair of teats on the side of the peritoneal cavity that has not received any injection.
   NOTE: Oxytocin is a small molecule that can be absorbed quickly and has a short half-life ^14,15. Milking should start within a minute after injection.
3. Hold the dam by gently scruffing her neck area with the non-dominant hand.
4. Manually express milk from a teat by squeezing the mammary tissue from the base to the tip of the teat until a bead of milk is visible.
5. Turn on the vacuum device by switching on the power button.
6. Start milking by gently applying the inverted P200 pipette tip to a teat. Use a gentle pulling motion to help draw out milk (Figure 1C).
7. Repeat steps 4.4–4.6 on all teats until a sufficient volume of milk is collected.
   NOTE: During milking, milk can accidentally be drawn into the 18 G hypodermic needle or PVC tubing of the vacuum line. This can easily be remedied by changing the needle or PVC tubing before milking the next animal.

5. Recovery of the dam after milk collection

1. Return the anesthetized dam to the cage with her pups and keep warm. Monitor the animal until she regains consciousness.
   NOTE: The dam may refuse to nurse again if she is allowed to wake up by herself in a separate cage before returning to her pups. This is more common when milk collection takes place during early lactation (PD2).
2. Let the dams move away from her pups to fully recover when she starts to wake up; this is a normal behavior.
   NOTE: We observe dams return to their pups periodically and return to nursing fully after a few hours.

6. Milk analysis

1. Use a commercial lactose assay kit (see Table of Materials) to measure the milk lactose. Dilute the milk 75x with sterile water and perform the assay according to the kit protocol.
2. Measure the milk fat using the creamatocrit method demonstrated in other publications^16,17.
3. Estimate the milk protein by performing the Bradford protein assay¹⁸ with bovine serum albumin as the standard¹⁹. Before performing the assay, centrifuge the milk at 4,000 × g for 15 min at 4 °C. Extract the infranatant by pipetting and dilute it to 1:10 with sterile water before the assay.
4. Perform polyacrylamide gel electrophoresis (PAGE); separate 10 µg of protein on a precast 4%–12% Bis-Tris gel as instructed by the manufacturer. Transfer the proteins onto polyvinylidene fluoride (PVDF) membranes and stain them with Coomassie Blue R-250.

Results

Using this technique, we have successfully sampled sufficient milk at different times during a 21 day lactation. We consider the day of delivery to be post-delivery day 1 (PD1), and in this work, we sampled milk on PD2 (early lactation) and PD10 (peak lactation), n = 4 each. We comfortably obtained 0.1–0.5 mL of milk from each dam (Figure 1D). To validate this method, we measured the gross macronutrient composition of these milk samples and compared the differences between milk collected at early and peak lactation using Student’s t-test.

We found that milk lactose concentration averaged at 12.3 ± 0.6 g/L on PD2 and 12.0 ± 0.9 g/L on PD10 (Table 1). Milk fat percentage was 26.0 ± 1.0% and 33.4 ± 2.5% on PD2 and PD10, respectively. The fat concentration was 174.4 ± 7.1 g/L on PD2 and significantly higher, 224.4 ± 17.1 g/L on PD10 (Table 1). The average protein contents were 70.1 ± 5.2 g/L on PD2 and 75.1 ± 2.6 g/L on PD10 (Table 1). Although the total milk protein content was not different between PD2 and PD10, the relative abundance of individual milk protein did differ. Specifically, αS1-casein and whey acidic protein were higher in PD10 milk than in PD2 milk (p < 0.05), while β-casein (p < 0.05) showed higher abundance in PD2 than PD10 (Figure 2).

Figure 2: Milk protein analysis. (A) Coomassie Blue-stained PAGE analysis of milk collected at PD2 and PD10. Protein identification was done as described previously¹⁴. (B) Individual protein signals were analyzed in ImageJ and normalized to the stained total protein. Results were expressed as the average with standard error bars and individual data points as dots. The average abundance of each protein on PD2 and PD10 was compared using a two-tailed Student’s t-test. *p < 0.05, n = 4. Abbreviations: PAGE = polyacrylamide gel electrophoresis; PDn = post-delivery day n. Please click here to view a larger version of this figure.

		Average	SEM	p-value
Lactose (g/L)	PD2	12.3	0.6	0.75
	PD10	12.0	0.9
Fat (g/L)	PD2	174.4	7.1	0.04
	PD10	224.4	17.1
Protein (g/L)	PD2	70.1	5.2	0.42
	PD10	75.1	2.6

Table 1: Macronutrient composition of milk collected at PD2 and PD10. Results, expressed as the average and standard error of the mean (SEM), were analyzed to examine the effects of day of lactation on milk composition using a two-tailed Student’s t-test, n = 4.

Discussion

We demonstrated a modified milking technique that can be performed by one person using a device that can be easily assembled and portable. Traditionally, rodent milk is collected using capillary tubes^10,17, glass Pasteur pipettes¹², or micropipettes¹¹. These methods require one person to restrain the dam and manually express milk and another to collect milk. This can sometimes be challenging, especially when milk sampling occurs during weekends or holidays. Milk collected by these methods must be transferred into sterile storage tubes. Since rodent milk is relatively high in fat (as shown here and in other studies^20,21), and therefore viscous, the process of transferring milk likely results in a loss of milk volume, limiting amounts for analyses. The device described here can be assembled in under 5 min with tools available in the laboratory. It can be easily used by one person to sample sufficient milk volume within 10 min. The continuous gentle suction approach associated with this method alleviates the issue of sample loss due to transferring since milk is constantly pulled straight into a collecting tube. Minimizing sample loss is particularly important when collecting milk as early as PD2 since milk yield is known to be relatively low at this stage¹⁰. Moreover, the same device and protocol were used to successfully collect milk from another mouse strain (FVB/N background), suggesting that this method likely works equally well across mouse strains.

In this study, we also observed that the time of separation may differ depending on the stage of lactation. Previous studies report that dams should be separated from their pups for at least 2 h and up to 16 h before milk collection to obtain sufficient milk volume^11,13. When employing this method, we found that separating dams from their pups for 2 h at early lactation (PD2) and for only 30–45 min at peak lactation (PD10) was sufficient to obtain an adequate volume of milk. Since it takes additional time for the dam to recover from anesthesia, minimizing the duration of separation not only reduces sampling time but also lessens the potential impact it may pose on the dam and her pups.

Additionally, we found that collecting milk as early as PD2 may negatively affect the subsequent growth of the pups due to the long separation and recovery time of the dam, leading to a period of insufficient feeding. This was not observed when milk was sampled at peak lactation (PD10). Therefore, if an investigator is interested in sampling milk at different time points during the 21 day lactation period, it is advisable to sample early lactation milk from a separate set of dams. Milk collection, however, can be performed on the same dam multiple times during peak and late lactation for longitudinal studies.

Alternatively, researchers may opt for other anesthetic agents to minimize the recovery time of the dam. For instance, isoflurane inhalation may help reduce the recovery time since animals typically recover quickly upon the discontinuation of isoflurane²². Although this choice of anesthesia is performed by many^13,23, it is reported that isoflurane inhalation may negatively affect the volume of milk collected¹³. Additionally, isoflurane inhalation requires appropriate apparatus to ensure precise control of dosage. Research on the effects of anesthetic agents such as ketamine/xylazine and isoflurane on lactation and milk composition and their potential transfer to breast milk is limited^24,25,26. It is, therefore, advised that data interpretation should be carefully performed when comparing studies using different choices of anesthesia. Another alternative is to sample milk without anesthesia, as described by other researchers^14,27,28,29. This procedure, however, can potentially cause excess stress to the dam and require an additional person to restrain the animal.

This study also showed that macronutrient assays could be performed using the milk samples, with a surplus remaining for more procedures such as omics analyses. Milk lactose, protein, and fat levels were measured and compared with published data to validate the described method. Milk lactose concentration accounted for ~1.2% of the milk component, which is similar to that of mice on a control diet reported in other studies^30,31,32. Mouse milk was confirmed to have a high percentage of fat, ranging from 26% to 33%, as reported previously^20,30. The total protein concentration reported here was similar to that reported by Chen et al.³⁰ but lower than that reported in other studies^31,32. In terms of changes from early to peak lactation, we found that lactose and protein levels remained relatively constant, but milk fat percentage increased from early to peak lactation. This is in disagreement with Ragueneau³², but the discrepancy may be due to differences in sample size. Although total protein concentration was similar between early and peak lactation, we found that individual protein abundance (αS1-casein, β-casein, and whey acidic protein) varied, presumably reflecting changes in the needs of the neonates at different developmental stages.

Although this technique does not allow quantitation of milk production, this can be calculated using the index of performance (IOP) proposed by Falconer in 1947³³. Milk production greatly depends on litter size and the day of sampling. This index can be used as a measurement of milk production accounting for litter size differences. For instance, the entire litter can be weighed on PD10 (when the dam is separated for milk sampling). The IOP of a dam is then calculated by dividing the weight of her litter by the mean weight of litters of the same size on PD10. In conclusion, we demonstrate an efficient single-person technique for milk sampling from laboratory mice using a device that can be easily assembled with relatively inexpensive and readily available equipment. This technique allows consistent collection of sufficient amounts of milk for most analyses.

Materials

Name	Company	Catalog Number	Comments
1 mL cryogenic vials	Corning	430658	Preferred for early milk collection
1 mL syringe	Exel	26048
2 mL microcentrifuge collecting tube	Fisher Scientific	02-681-344	Preferred for peak and late milk collection
18 G hypodermic needle	BD PrecisionGlide	305195
27 G hypodermic needle	BD PrecisionGlide	305109
Alcohol prep wipe	Honeywell	154818
Bio-Rad Protein Assay Dye Reagent Concentrate	Bio-Rad	5000006
Bovine Serum Albumin	Sigma-Aldrich	A2153
Immobilon -P PVDF Membrane	Immobilon	IPVH00010
Ketamine	Dechra	1000001250	Purchased via Veternarian Office
Laboratory labeling tape	VWR International	89097
Lactose Assay Kit	Sigma-Aldrich	MAK487
Mini Barbed polypropylene fittings	Cole-Parmer Instrument Company	6365-90
Mouse diet	Lab Diet	5015
NuPAGE Bis-Tris Mini Protein Gels, 4–12%	Invitrogen	NP0335BOX
Optixcare Eye Lube	Aventix	N/A	Purchased via Veternarian Office
Oxytocin	Bimeda	1OXY015	Purchased via Veternarian Office
P200 pipette tips	Gilson	F1733001
PVC tubing: 1/8'' ID	VWR International	MFLX07407-75
PVC tubing: 5/32'' ID	VWR International	MFLX07407-77
Septum stopper	Chemglass Life Sciences	CG-3022-91
Vacuum pump	Drummond Scientific Co.	P-103635
Xylazine	Akorn Animal Health	NDC 59399-110-20	Purchased via Veternarian Office