Repix is designed with few and simple components.

A. Repix has only two principal components: payload and docking modules. Probe cassettes and posts are constructed in aluminum, shown front (top) and back (bottom). A model of a Neuropixels probe 2.0 multi-shank is shown for scale. Dashed circles highlight features that are discussed in the text (“Edge”, “Feet”, “Screw”). B. Chronic implantation with Neuropixels 1.0 probes. The system is assembled first by fixing the probe to the cassette. The cassette is held by a 3D-printed connector, top. The cassette and probe can then be moved independently of the posts, which lock onto the cassette with M1 screws. The cover and headstage holder complete the system. C. The procedure to implant the probe. The connecter is attached to the channels of the cassette and secured with screws, second panel. Posts are attached from the cranial side and secured with screws, third panel. Finally, once the probe is in place, the feet of the posts are cemented onto the skull of the animal and the connector can be loosened, fourth panel. D. Fully assembled implant with the Neuropixels 1.0 cover and headstage holder. E. The headstage can be attached with a headstage holder that fixes into the slot previously used for the connector. There are versions for Neuropixels 1.0, left, and the further miniaturized Neuropixels 2.0, right. F. A cover is used to further protect the shanks. The 2.0 cover consists of a base and a box that snaps into place, with access to the 4-pin connector of the headstage. CAD files for all components are available along with a detailed surgical protocol. G. The system in place in a mouse (CD-1, age: 10 weeks) with a 2.0 cover.

Repix has been implemented across multiple labs and users.

A. The total number of procedures across all users in implantations, data collection, and explantation (success: purple, failure: white). These correspond to 97%, 89%, and 83% success rate, respectively. B. Fifteen of sixteen users responded to a survey of their use of Repix. Users had high individual success rates in implanting Neuropixels probes using Repix. The length of the bars corresponds to the number of procedures undertaken (same order as in A), colored for success and white for failures. Users who trained others are connected by lines. Most, but not all, users also had high individual success rates in data collection and explantation of Neuropixels probes using Repix.

Repix can be implanted for up to a year with stable unit yield.

A. The summary of recording sessions in 27 mice implanted with Repix. These implantations were in the hippocampus (red), entorhinal cortex (orange), amygdala (green), and V1 (blue). The maximum duration was 364 days. B. Individual implants maintained stable unit yield over weeks and months, both in total units (left) and good units (middle). These units had a stable waveform amplitude over time (right). Thin lines are logarithmic fits to the recording data from the individual and thick lines represent logarithmic fit to binned average across animals. Note the logarithmic time scale.

Repix allows reuse of Neuropixels probes.

A. Across all users, 30% of implantations were done with reused probes. B. Nine of fifteen users opted to use Repix’s feature to reuse probes. Seven of fourteen users supplied data on reusing their probes. Colored bars indicate the number of implantations undertaken with a reused probe. C. The initial-day yield of good units in the amygdala did not decline as the same probe was reused multiple times. Data points from individual probes are connected by lines. Circles are colored by their surgical date rank, early to late implants from light to dark. D. Left: New (gray) and reused (orange) probes implanted in the entorhinal cortex had similar unit yield, defined as the peak number of units in a single session, left axis. Equally, waveform amplitudes were similar, right axis. The dots represent data from individual implants. The mean of implantations is represented in the circles with s.e.m. in the shading. Right: Exponential fits show no significant difference in good unit yield stability. Waveform amplitude was similarly stable. Same color conventions as in left panel. A significantly quicker drop-off in total units was observed (Fig S3).

Repix can be used in many different experimental paradigms.

A. Unrestrained implanted mice move naturally. Implanted mice (purple) showed similar movement dynamics to non-implanted mice (gray) across the spectrum of slow (low frequency) to fast (high frequency) head movements during social interactions, both in pitch (main panel) and roll (inset). B. Repix can be used with any Neuropixels probe type. Histogram of good unit yields by probe type, Neuropixels 1.0 in purple, Neuropixels 2.0 single shank in dark gray, Neuropixels 2.0 multi-shank in light gray (mean yield in legend). While the users reported good yields across areas and probes, each user tended to use a single probe type in each brain area, making direct comparisons of yield by type within an area difficult. C. Left: Repix parts for Neuronexus probes. Right: Success rates of each experiment stage, including reuse of probes, using the modified design. D. and E. Repix is fully compatible with chronic recording in rats. Stability and yield from mPFC/ACC (D, n=5 rats) and amygdala recordings (E, n=6 rats). Same conventions as Fig 3. E. Rats injected with dexamethasone (green, n=3) had improved yield and stability compared to non-injected animals (gray, n=3).

Users quickly gain proficiency.

Top: Each procedure (attempted/planned implant) was categorized as either: Practice (gray), Implantation failed (red), Data Collection (light green), Explantation failed (green) or Full procedure (dark green), for nine users who supplied detailed surgical records. *User VP implanted Repix without intention to explant, and therefore Full Procedure includes only Implantation and Data Collection (Dark green, dotted). Bottom: Proficiency was defined as three sequential successful Full procedures, and the median number of procedures to reach this point is plotted (green circle, IQR in shading).

Overview of researchers and laboratories.

Repix was shared with sixteen users. Fifteen responded to our census and seven provided electrophysiological yield data (green).

Questionnaire used for the census

Users were sent the above questionnaire to supply details of each procedure attempted and its outcome, along with meta-data.

Yield summary

A: The mean yield per implantation, left, and max yield per implantation, right. B: Three estimates of the stability of the yield: Left, the mean yield divided by the first recording yield, left; middle, the mean yield divided by the max yield at recorded on that implantation; right, the exponential fit to the yield curves over time. Total count, including MUA, only good units, and their associated amplitude, left, middle and right in each panel. Color conventions as in Fig 3.

Reuse summary

No significant difference between new (gray) and reused (colored) probes in the entorhinal cortex. The mean yield per implantation, left, and max yield per implantation, middle. The exponential fit to the yield curves shows a significant difference between new and reused probes only in MUA+good units (ie. total units) and a non-significant difference in Good units. Waveform amplitudes are unchanged.