Work Description

Title: Dataset for D1 and D2 Comparisons in Mouse Self-Stimulation Open Access Deposited

http://creativecommons.org/licenses/by-nc/4.0/
Attribute Value
Methodology
  • Two independent self-stimulation procedures were used to allow mice to earn brief NAc laser stimulations as a reward. First, a spout-touch self-stimulation procedure allowed mice to earn brief 1-sec laser pulses in NAc each time they touched a particular metal spout (which we will call laser-spout) that protruded into the chamber. Another spout was also present, but earned nothing when touched, and served as a control stimulus for comparison. Second, a separate place-based self-stimulation task allowed mice to earn series of laser pulses by entering a particular corner of a 4-corner chamber and remaining there; this task was based on the original Olds and Milner procedure that discovered deep brain self-stimulation when a rat went to a particular location to activate its electrode. After the location task, some mice were also retested on the spout task to reconfirm their initial results.
Description
  • The nucleus accumbens (NAc) contains multiple subpopulations of medium spiny neurons (MSNs): one subpopulation expresses D1-type dopamine receptors, another expresses D2-type receptors, and a third expresses both. The relative roles in NAc of D1 neurons versus D2 neurons in appetitive motivation were assessed here. Specifically, we asked whether D1-Cre mice or D2-Cre mice would instrumentally seek optogenetic self-stimulation of those respective subpopulations in NAc, or instead avoid NAc laser stimulation.

  • Some statistical forms have been exported from SPSS for the purposes of accessibility to viewers. Please see the "readme" text for descriptions of each individual excel file.
Creator
Depositor
  • shinok@umich.edu
Contact information
Discipline
Funding agency
  • National Institutes of Health (NIH)
Keyword
Resource type
Last modified
  • 11/05/2019
Published
  • 04/04/2018
Language
DOI
  • https://doi.org/10.7302/Z2N014RK
License
To Cite this Work:
Cole, S. Dataset for D1 and D2 Comparisons in Mouse Self-Stimulation [Data set], (2018). University of Michigan - Deep Blue. https://doi.org/10.7302/Z2N014RK

Relationships

Files (Count: 6; Size: 2.67 MB)

Date: 12 November, 2018

Title: Dataset for D1 and D2 Comparisons in Mouse Self-Stimulation
Authors: Shannon L Cole
Contact: Shannon Cole, shinok@umich.edu
Funding: National Institutes of Health (NIH)
------------------------

Research Overview:

The nucleus accumbens (NAc) contains multiple subpopulations of medium spiny neurons (MSNs): one subpopulation expresses D1-type dopamine receptors, another expresses D2-type receptors, and a third expresses both. The relative roles in NAc of D1 neurons versus D2 neurons in appetitive motivation were assessed here. Specifically, we asked whether D1-Cre mice or D2-Cre mice would instrumentally seek optogenetic self-stimulation of those respective subpopulations in NAc, or instead avoid NAc laser stimulation.

Mice received Cre-targeted channelrhodopsin (ChR2) virus and optic fibers in NAc. Subsequently, mice could earn brief NAc laser illuminations by actively touching a metal spout in one task, or by going to a particular location in a separate task. Results indicated that D1 neuronal excitation in NAc supported intense self-stimulation of hundreds to thousands of spout-touches per half-hour session. D1 Cre mice also sought out locations that delivered NAc laser to excite D1 MSNs. By comparison, D2 ChR2 mice showed lower but still positive levels of self-stimulation in the spout-touch task, earning dozens to hundreds of NAc laser illuminations. However, in the location task, most D2 mice either ignored laser or instead gradually avoided the laser location.

Brain-wide measures indicated that D1 and D2 stimulations in NAc recruited heavily overlapping patterns of Fos activation in distant limbic structures. These results confirm that excitation of D1 MSNs in NAc supports strong incentive motivation to self-stimulate, and suggest that excitation of D2 neurons in NAc produces a more plastic or ambivalent motivation, which can shift between positive versus negative valence.

Some statistical forms have been exported from SPSS for the purposes of accessibility to viewers. Please see the "Files" section for descriptions of each individual excel file and descriptions for notation used.
------------------------

Methodology:

Two independent self-stimulation procedures were used to allow mice to earn brief NAc laser stimulations as reward. First, a spout-touch self-stimulation procedure allowed mice to earn brief 1-sec laser pulses in NAc each time they touched a particular metal spout (which we will call laser-spout) that protruded into the chamber. Another spout was also present, but earned nothing when touched, and served as a control stimulus for comparison.

Second, a separate place-based self-stimulation task allowed mice to earn series of laser pulses by entering a particular corner of a 4-corner chamber and remaining there; this task was based on the original Olds and Milner procedure that discovered deep brain self-stimulation when a rat went to a particular location to activate its electrode (1).

After the location task, some mice were also retested on the spout task to reconfirm their initial results.

Subjects
BAC transgenic mice on a C57Bl/6 background (n= 59) were obtained from NINDS/GENSAT (www.gensat.org) from Rockefeller University/NIH/NIMH, and maintained on a 12-hour reverse light-dark cycle with food and water ad libitum. These included 33 D1-Cre mice (12 male, 21 female; strain: B6.FVB (Cg)-Tg (Drd1a-cre) EY262Gsat/Mmucd), and 26 D2-Cre mice (12 male, 14 female; strain: B6.FVB (Cg)-Tg (Drd2-cre) ER44Gsat/Mmucd). D1-Cre females and males, and D2-Cre females and males were randomly assigned to receive NAc infection with either a channelrhodopsin virus (AAV5-DIO-ChR2-EYFP) to be an optogenetic group, or an optically-inactive virus to be an EYFP-only control group that lacked the ChR2 gene (AAV5-DIO-EYFP). This created four Cre/Virus groups for all following experiments 1) ChR2 D1-Cre [n=14 total (4 male, 10 female)]; 2) ChR2 D2-Cre [n=14total (7 male, 7 female)]; 3) EYFP D1-Cre control [n=19 total: (8 male, 11 female)]; 4) EYFP D2-Cre control [n=12 total: (5 male, 7 female)].

Male and female mice were housed separately, and test chambers were cleaned after each mouse was tested to avoid pheromone contamination. All animal protocols were in accordance with the National Institutes of Health Guide for Care and use of Laboratory Animals and approved by the University Committee on the Use and Care of Animals at the University of Michigan.

Viral Vectors
A DIO Cre-dependent ChR2 Adeno-associated virus (AAV) was used to infect Cre-expressing cells (vectors-double loxP-flanked inverted (DIO) - channelrhodopsin 2 (ChR2) - enhanced yellow fluorescent protein (EYFP) (AAV5-DIO-ChR2-EYFP; purchased from the University of North Carolina Vector Core with MTA by courtesy of Karl Deisseroth and Stanford University).

Surgery
Mice were anesthetized with isoflurane gas (4-5% induction, 1-2% maintenance), placed in a stereotaxic instrument (David Kopf Instruments), and the skull surface was exposed. Bilateral microinjections of virus (0.5 μl per side; 0.1 μl/min) into NAc were targeted at the medial shell and medial core. Either ChR2 virus (AAV5-DIO-ChR2-EYFP) or an optically-inactive control virus (AAV5-DIO-EYFP) (0.5 μl) was delivered via 28 gauge syringe over 5 min, and left unmoved for 10 minutes to allow for viral diffusion. Stereotaxic coordinates for virus microinjections centered around AP +1.42 to +1.32; ML +/- 1.5; DV – 4.78; injectors were angled at 12.29 lateral degrees to avoid ventricles and permit space for bilateral fiber implants. NAc sites were staggered slightly across individual mice to nearly fill the entire medial shell as a group, and include some penetrations in core (AP coordinates range from +1.42 to +1.32), but within a mouse bilateral sites were kept as symmetrically identical as possible. In the same surgery, optic fibers (6 mm long; 0.220μm core; confirmed to exceed 85% light efficiency prior to surgery) were bilaterally implanted in NAc approximately 0.3 mm above each site of virus injection (AP + 1.42 to +1.32; ML +/- 1.5; DV – 4.48; 12.29 degree lateral angle).

Optic fibers were anchored to the skull using surgical screws and dental acrylic. Mice were allowed at least 4 weeks after surgery for incubation and virus expression before behavioral tests began.

Experiment 1: Laser self-administration-spout touch tasks
NAc self-stimulation was tested first using a spout-touch self-administration task, in which active touches of a designated empty metal drinking-spout could earn phasic 1-sec illuminations of NAc laser.

Optic fibers were attached through an FC/PC adaptor to a 473 nm blue DPSS laser (OEM Laser Systems). Two empty metal spouts (lickometer touch-capacitive detectors) protruded through the wall of the 8x10x5cm chamber (MedAssociates Inc.), placed approximately 5 cm apart (see S1 File for example). Active touches of one arbitrarily-designated spout (laser spout) delivered a brief 1-sec bin of laser illumination activated by an Arduino control board (Arduino Hardware), and accompanied by a distinctive auditory cue that served as a sensory label for the laser-delivering spout (either white noise or tone, 1 sec).

Touching the other control spout produced no laser, but did produce the different auditory cue as a distinctive marker, and touches on it served merely as a control measure of generalization, exploration and general motor activity (no-laser spout). In an initial screening wave of mice (n = 6 D1 ChR2; n=5 D2 ChR2; n=7 D1 EYFP & n=6 D2 EYFP), an initial screening compared the relative effectiveness of constant laser versus pulsed 25 Hz laser in 1-sec illumination bins during 30 min sessions. Both types of laser stimulation were compared at three different laser intensities: 0.1, 1.0 and 10 mW.

The same spout location delivered laser on all days (paired either with white noise or tone). The other spout remained the control inactive spout on all days (paired with the other sound; spout/noise-tone assignments were balanced across mice). Optic fiber light transmission at the end of the output optic cable was confirmed each day (Laser Check Photometer, Clairvoyance Inc.), and cranial fiber implants had been tested for 85% efficacy prior to surgery.

Individual D1 ChR2 mice or D2 ChR2 mice were excluded from being considered self-stimulators if they failed to meet both criteria of 1) at least 1 session of 20 contacts on the laser-delivering spout and 2) a 2:1 ratio of laser-paired versus non-laser paired contacts on at least one test day.

Constant 1-sec illumination vs pulsed 25 Hz stimulation
Pulsed laser stimulation at a particular frequency (e.g. 25 Hz) is often used in optogenetic studies to drive neuronal firing at the same frequency. By contrast, constant laser illumination at low-intensity (e.g., 1-2 mW) has been suggested to avoid driving neuronal firing at any particular artificial frequency, and instead to promote striatal endogenous firing patterns (e.g., 10% – 30 %) without significantly altering patterns of natural wave-form potentials (21, 37-40).

To compare efficacy of pulsed vs constant NAc stimulation, mice received 1-sec bins at each of the 3 illumination intensities of either constant or pulsed laser (order of pulsed/constant conditions was counterbalanced across mice).

Dose-response comparison (0.1, 1, 10 mW intensities)
To compare the relative effectiveness of different intensities of laser illumination, laser intensity was changed consecutively between 3 sessions at each of the 3 intensity levels: 0.1mW, 1mW, 10mW (either at constant illumination or at 25 Hz, both 1-sec duration; order counter-balanced across mice).

Location-tracking 9-day groups
The initial screening test results identified the middle 1.0 mW intensity and the constant-illumination 1-sec parameters as suitable for producing moderate levels of NAc self-stimulation. However, screening results also showed within-group variability for each condition that was higher than optimal statistical comparison, leading us to follow up with a more extended 9-day test with a single laser setting. This extended test with constant-illumination, 1.0 mW intensity, and 1-sec bin parameters was intended to stabilize self-stimulation rates and potentially reduce variability (n = 9 D1 ChR2 mice ; n=9 D2 ChR2; n=6 D1 EYFP and n=6 D2 EYFP). Additionally, we wished to assess whether self-stimulations truly motivated in the sense of being instrumental actions that were directed flexibly and specifically aimed at obtaining NAc ChR2 laser excitations, or instead if spout-touches were merely being stamped-in rigid stimulus-response habits or simply repeated as a mere motor reaction to an immediately prior NAc stimulation.

Therefore, during the 9-daily sessions we shifted the location of the laser spout three times (a shift every 3 days) to test whether mice would flexibly track the source of NAc laser as spouts moved. . The location of the two spouts were fixed across the first three days of testing (Days 1-3; active/inactive locations were counter-balanced across mice). On Day 4, the active laser spout and inactive spout were both moved to new positions on the opposite wall, and then kept stable over Days 4-6. On Day 7, both spouts were moved back to the original wall, but reversed from their original positions on days 1-3, so that the laser spout now occupied the former no-laser spout position and vice versa, and kept in their new positions over Days 7-9. This presented the mouse with a 3rd new location for the active spout, which was exactly opposite to its original location.

Laser-extinction test
Finally, on the 10th day, an extinction session with no laser was given to further assess if self-stimulation behavior became habitual or aimed at conditioned reinforcement, or instead remained flexible and dependent on NAc laser activation. In the extinction session, the laser reinforcement was discontinued, though touching each spout still produced its associated auditory cue.

Experiment 2: Location-based self-stimulation task
A second location-based self-stimulation task was conducted subsequently, which allowed NAc laser stimulations to be earned more passively by simply entering or remaining in a particular corner location in a 4-corner chamber.

The center of the 90x90x106cm chamber was occluded by a large cylinder (20cm diameter plexiglass), so the mouse could circumnavigate only along the outer periphery and among the four corners of the chamber, and the floor of the chamber also contained bedding. Within the laser-delivering corner, any movement triggered an infrared motion detector that delivered a 1-sec laser constant illumination (1mW, constant) per movement during a 30-min session. Each corner had its own motion detector (Visonic), and MATLAB software was utilized to compile entries and time spent within each of four corners.

One corner was arbitrarily designated as laser-delivering each day (corner assignment balanced across mice; corner changed each day for a particular mouse). Laser stimulations were earned on entry and by every further movement detected while the mouse remained within that corner. Laser immediately ceased when the mouse left the designated corner. Entries and time spent in the laser-corner was monitored. Entries and time spent in the other three corners were also monitored but did not produce laser illuminations. On the second test day, the active corner designation was shifted to the corner opposite the Day 1 laser-corner. On the third day, the active corner designation was arbitrarily shifted to a new location in one of the remaining two corners.

One reason this location-based procedure was added is that it can assess laser avoidance of the laser-corner, as well as preference, as avoidance would be evident by a mouse’s not entering or more quickly leaving the laser-delivering corner compared to alternative corners. Behavior was also videotaped each day for subsequent video analysis, and scored for seconds duration engaged in burrowing (submerging head and using bilateral forepaw movements to throw bedding backwards), defensive treading (throwing bedding forward via alternating unilateral forepaw movements), rearing (elevating body and head together on hindpaws so that forepaws rose >1cm above floor), locomotion (seconds of continuous forward movement), and immobility.

Histological analyses of virus expression, local Fos Plumes, and distant Fos activations
Immediately before euthanasia, a standardized dose of laser stimulations was passively delivered to all mice in order to 1) generate local Fos plumes of neuronal Fos activation immediately surrounding the fiber tip in NAc, and 2) to potentially recruit distant Fos activations in other brain structures that would reflect functional connectivity patterns for D1 vs D2 circuitry. Beginning 90 minutes prior to euthanasia and perfusion, each mouse was put into a self-administration spout chamber and given 1s bursts of 1mW laser stimulation every 10 seconds for 90 min.

Laser stimulation was not contingent on any behavior prior to perfusion, so as to ensure equal laser exposure for D1-Cre and D2-Cre mice. All mice were then deeply anesthetized with an overdose of sodium pentobarbital, transcardially perfused, and brains were removed and analyzed for Fos plumes, and for distant Fos recruited in other limbic structures.

Briefly, brains were stored in 4% paraformaldehyde for 1 day post-perfusion, and then soaked in a 30% sucrose solution for 2 days prior to slicing. 40 micron slices were blocked in 5% normal donkey serum/0.2% Triton-X solution for 30 min before being incubated for 24h in a polyclonal rabbit anti-c-fos IgG primary antibody (Santa Cruz Biotechnology; 1:1000 dilution; lot #K0415, RRID: AB_2106783), followed 1 day later by Alexa Fluor 594 donkey anti-rabbit IgG secondary antibody (Life Technologies; 3:1000 dilution; lot #1668652, RRID: AB_141637). Sections were mounted, air-dried, and cover slipped with ProLong Gold anti-fade re-agent (Invitrogen).

Fos plumes, or local neurons expressing Fos surrounding an optic fiber tip, were counted using a grid with 8 arms emanating from the fiber tip, each arm containing consecutive 50-micron boxes, which were overlain until Fos decreased to baseline levels or reached the medial hemisphere. This analysis permitted calculation of the relative increase or decrease within a set area from the fiber optic tip, yielding a Fos density plume. Distributed Fos was calculated by counting Fos within selected regions within a 600x800-micron square. Viral infection was also measured by placing a similar grid overlay, and measuring EYFP fluorescence in 50-micron increments until levels fell to baseline levels.

Statistics
Non-parametric Friedman’s ANOVA was used for initial within-subject repeated measures, and Kruskal-Wallis ANOVAs for between-subject comparisons. If initial analyses were significant, additional Mann-Whitney or Wilcoxon tests were used as appropriate for subsequent paired comparisons. For all analyses, the significance level was set at p = .05, two-tailed. Effect sizes for pairwise comparisons were calculated using the following formula: r=Z/√(N1+N2).

References:
1.
------------------------

Files:

The following is a summary of the individual excel files provided and descriptions for notation used. Some statistical forms have been exported from SPSS for the purposes of
accessibility to viewers.

1) "Figures 1_2_-Spout Self-Stimualtion.xlsx"

This file contains the data used for Figures 1 and 2, where mice were tested across 9 different days and 3 different
bottle spout positions in order to determine if mice would work for D1 or D2 neuron stimulation and if they would
seek laser depolarizations.

Tab1-Individual Self-Stimulation 9D: This tab presents the overall test averages, daily averages, and the
individual daily values. Day 9 vs extinction values are shown at the right end of the sheet. Laser-paired vs non-paired
refers to the laser-paired bottle spout or non-paired spout able to be contacted.

Sex is encoded as "1"="Male" and "2"="Female".
D1_D2_Code is "1"=D1 ChR2, "2"=D2 ChR2, and "3"="EYFP Control"

Day refers to Test Days 1-9. "Position"refers to the configuration of bottle spouts in 3 configurations.

"Laser-Paired" and "Non-paired" refer to laser-paired vs non-laser paired spouts, respectively.

Tab3-Timecourse TD1, TD4, and TD7: The following data are the 1 minute binned data points for Test Days 1, Test Day 4,
and Test Day 7. On these 3 days, the bottle spout configuration was altered to determine if the mice were motivated
to seek out laser-paired spouts.

D1 ChR2 Condition, D2 ChR2 Condition or EYFP Control Conditions are designated in the top left hand corner and Test Day 1, 4, and 7 are
marked by yellow highlight.

2) "Figure_3_Location_Self-Stimulation.xlsx"

This file contains data used for Figure 3. Mice were given free access to a chamber where 4 with infrared break
beams in each corner. Each mouse was run for 3 test days and data was analyzed for beam breaks, time in each corner,
and locomotion. Each day, the laser was moved to a new test corner to determine if the mouse would seek laser-paired
locations.

Tab1-Operant Place Preference: Individual values for beam breaks ("Breaks) and time spent ("Time") in corner organized
by day (left) and averages shown (right).

Laser-paired refers to the corner paired with laser-stimulation; right refers to the corner immediately right of
laser-paired corner; left refers to corners left of the laser-paired corner, and opposite is across from the laser
corner on particular test days.

Virus_Condition is "1"=D1 ChR2, "2"=D2 ChR2, and "3"="EYFP Control"
Day is in the order of test days.
"Breaks" or "Time" is divided by individual corner activations.

Tab2-Locomotion_Distance_2SecSample: This tab contains individual measurements of distances traveled in 2 second bins following laser activation.
5 bins per test day were taken per animal (15 bins total). Top values give individual bin measurements and bottom values represent averages
per individual test day.

Virus conditions refer to D1 ChR2 (1) or D2 ChR2 (2) virus conditions, respectively.

Tab3-Locomotion Data:

Data provided in this tab are from offline video scoring to determine if exposure to laser-stimulations produces
alterations in locomotion. Given are the total durations of time in locomotion, immobile, or grooming.

D1 ChR2, D2 ChR2 or EYFP control conditions are highlighted in yellow.

3) "Figure_4_Fiber Placements vs Max Stimulation.xlsx"

This file contains data used in Figure 4. The fiber optic placements were arranged along medial-lateral coordinates, dorsal-ventral coordinates, and rostral-caudal coordinates.

For Medial-Lateral Coordinates: Placements ranging (in mm) from 0-0.25=1; 0.26-0.5=2; 0.51-0.75=3; 0.76-1.0=4.
For Dorsal-Ventral Coordinates: Placements ranging (in mm) from 4.0-4.24=1; 4.25-4.49=2; 4.5-4.74=3; 4.75-5.0=4.
For Rostral-Caudal Coordinates: Placements ranging (in mm) from 0.75-0.99; 1.0-1.24; 1.25-1.49; 1.5-1.75.

Max_stim refers to the maximal number of self-stimulations during one 30 minute trial at 1mW, constant stimulation.

D1_D2 Code designate D1 ChR2 (1) or D2 ChR2 (2) mice.

4) "Figure_5_Fos_Virus Plume and Self-Stimulation Behavior.xlsx"

This file contains 3 sections: a section showing maximum self-stimulation rates vs virus spread; a plot of the number of 50um boxes filled by EYFP coming from virus injection site;
a final section listing the number of cells counted for the density of Fos cells counted within the NAc.

Tab1-EYFP Expression vs Behavior-Here values are given for correlations of self-stimulation on the 9-day spoute self stimulation task and location-stimulation task
versus virus expression to determine if virus spread affects self-stimulation behavior.

D1_D2 conditions encode for D1 ChR2 (1) vs D2 ChR2 (2) conditions.
EYFP radius and diameter values are measured in mm.
Average_SS_Laser_Paired and Maximal_SS_Laser_Paired refers to the number of contacts on the laser spout across 9-days.
OPP_3_Day_Average_Break and OPP_3_Day_Average_Time give values for time spent or infrarred beam breaks within the laser-paired corner.
OPP_Day_3_Breaks and OPP_Day_3_Time give values for test day 3 only.

Tab2-Virus Plume Plotter-D1 ChR2 and D2 ChR2 virus expression values were calulated by using 8-radial arm overlays comprised of 50um boxes
and the number of boxes filled by EYFP expression were counted. The values for D1 ChR2 virus spread is designated in the top left corner
and D2 ChR2 values are designated in the bottom left corner.

Tab3-FosPlume_BoxCounts-D1 ChR2, D2 ChR2, EYFP control, and surgery naieve mice had Fos values calculated using an 8-radial arm overlay
comprised of 50um boxes. The values within each cell represent the number of Fos+ cells within each box.

Column A provides the direction from the optic fiber-implant site. The value next to each cardinal direction denotes the box.
E.g, NE1=the first box in the North East radial arm.

Blank boxes represent non-counts due to tissue damage or no tissue present in the overlay view frame.
Picture represents the specific image used for calculations.
Group denotes the specific surgical condition.

5) "Figure_6_Distant Fos.xlsx"

This section lists the values of Fos+ cells across multiple brain regions calculated by counting 600umx800um boxes within a hemisphere. Fos was induced by 1s bursts of
laser-stimulation every 10 seconds for 30 minutes.

Top sections represent the Fos+ neuron counts per each hemisphere in each brain region. Bottom sections show the total neurons counted
per animal.

Brain region abbreviations are: Prelimbic cortex (PrL), Infralimbic cortex (IL), Basolateral Amygdala (BLA), central nucleus of Amygdala (CeA),
ventral subiculum (vSub), rostral NAc Shell (rShell), rostral NAc core (rCore), caudal half of NAc medial shell (cShell), caudal NAc Core (cCore),
rostral ventral pallidum (rVP), caudal Ventral pallidum (cVP), lateral hypothalamus (LH), ventral tegmental area (VTA), substantia nigra (SN).

Virus Conditions:
1=D1 CHR2; 2=D2CHR2; 3=D1 EYFP; 4=D2 EYFP; 5=surgery naieve mice.

------------------------

Related publication(s):

Use and Access:
This data set is made available under a Creative Commons Attribution-NonCommercial 4.0 International license (CC BY-NC 4.0).

To Cite Data:
Cole, S.L (2018). Dataset for D1 and D2 Comparisons in Mouse Self-Stimulation [Data set]. University of Michigan Deep Blue Data Repository. https://doi.org/10.xxxx/xxxxxxxx

Download All Files (To download individual files, select them in the “Files” panel above)

Best for data sets < 3 GB. Downloads all files plus metadata into a zip file.

Files are ready   Download Data from Globus
Best for data sets > 3 GB. Globus is the platform Deep Blue Data uses to make large data sets available.   More about Globus